CSENDEPLRUUKBG
00

๐ŸŒ CZECHAI โ€” MASTER CATALOGUE OF THE ECOSYSTEM

One interconnected technology ecosystem that brings together proprietary models, GPU infrastructure, agents, data, applications, products and specific workflows into solutions ready for real-world deployment.

01

๐Ÿงญ MASTER STRUCTURE

I. Services, Products and Industry Solutions ๐Ÿงฐ โ€” what CzechAI offers and delivers to customers.
II. Technology, Product and Development Areas ๐Ÿš€ โ€” what solutions are built from and where the ecosystem is heading.
III. Claude Arsenal ๐Ÿงฌ โ€” a standalone internal catalogue of agents, swarms, MCP, skills, plugins, commands and operational rules.
IV. Verified Metrics, Data and Infrastructure ๐Ÿ“Š โ€” numerical data with measurement sources.
V. Main Public Platforms and Outputs ๐Ÿ”— โ€” a unified navigation hub.
VI. Master Rules ๐Ÿ“ โ€” methodology for completeness, deduplication and updates.
02

๐Ÿงฐ I. SERVICES, PRODUCTS AND INDUSTRY SOLUTIONS

1

๐Ÿง  Proprietary LLM & AI Laboratory

๐Ÿ–ฅ๏ธ Proprietary GPU and LLM Infrastructure โ€” we operate self-hosted models, a model gateway, intelligent routing, load balancing and automatic failover between proprietary and external models, giving clients performance, data control and operational independence.
๐Ÿงช Model Adaptation and Fine-Tuning โ€” we tailor models to Czech language, domain terminology, company documents, tool-calling and the client's specific processes, so that responses and actions reflect the organisation's real-world practice.
๐Ÿ“š RAG and Knowledge Systems โ€” we connect Qdrant, hybrid search, knowledge graphs, internal documents and verifiable sources into a proprietary Czech knowledge layer cz-rag, which delivers traceable answers over current data.
๐Ÿค– Agents and Orchestration โ€” we build specialised agents, MCP tools, multi-agent workflows, long-term memory and automatic task handoffs between roles, so that complex work runs like a coordinated digital team.
2

๐Ÿ—ƒ๏ธ Data Intelligence & Data Pipelines

๐Ÿ•ท๏ธ Data Collection and Enrichment โ€” we acquire data from websites, documents, registries and the client's internal systems, and handle scraping, OCR, normalisation, deduplication, matching and continuous updates.
๐Ÿ›๏ธ Public and Reference Sources โ€” we connect ARES, ฤŒรšZK, RรšIAN, ISIR, auctions, official notice boards, case law and other verifiable sources into a unified data layer suitable for analysis and automation.
๐Ÿ—บ๏ธ Geodata and Map Applications โ€” we create geocoding, risk and thematic layers, interactive maps, spatial analyses, 3D visualisations and location intelligence for territory-based decision-making.
๐Ÿ”„ ETL, API and Reporting โ€” we transform raw data into databases, APIs, dashboards, XLSX/CSV exports, alerting and downstream automations; this includes worldmonitor at monitor2.czechai.io for continuous global event monitoring.
3

๐Ÿ  Real Estate, Valuation & Investment

๐Ÿ“ˆ Price Map V26 โ€” estimates the value of a flat, house or plot from an address or listing URL, providing a price range, comparable offers and a reliability score; available at shadow.czechai.io.
๐Ÿ’Ž Below-Market Opportunities โ€” automatically identifies undervalued listings, direct owners, off-market opportunities and other investment signals that are easily overlooked in standard advertising.
๐Ÿ”Ž Reality Portal & Copilot โ€” combines listings, market data, expected yields, risks and property due diligence into a single working environment; the main platform is EstateX at estatex.club.
๐Ÿ”จ Auctions & Acquisition โ€” monitors auctions, compares opening and final prices against the market, and prepares materials for acquisition, investment assessment and rapid decision-making.
๐Ÿ“ก Lead Hub โ€” delivers investment leads on a turnkey basis: direct owners, long-running and repeated listings, motivated sellers, undervalued properties, auctions and insolvencies, with output to a spreadsheet, API or CRM.
๐ŸŒ Territorial Development Prediction Platform โ€” evaluates the probability of future land reclassification to building plots based on zoning plans, รšAP, ZรšR and รšSES; the solution is sold exclusively to a single client.
๐Ÿ“ฑ Essty / EstateX Mobile Application โ€” a real estate agent in your pocket, connecting listings, valuations, clients, notes and voice control into a single mobile work tool.
๐ŸŽฎ HรกdejCenu & Hypotown โ€” interactive applications for working with property prices and locations that educate users, capture relevant leads and simultaneously serve as marketing tools.
๐Ÿ˜๏ธ Residential Property Management โ€” a white-label application for homeowners' associations and property managers that connects resident communication, billing, fault reporting and tradesperson dispatch.
4

โš–๏ธ Legal AI & Document Intelligence

๐Ÿ“– AI Lawyer โ€” conducts legal research across regulations, case law and an organisation's internal knowledge base, returning answers with traceable sources for subsequent expert review.
๐Ÿ“„ Contracts & Filings โ€” analyses contracts, flags risky clauses, prepares redlines, templates, document drafts, procedural filings and a clear chronology of a dispute.
โš”๏ธ Litigation Support โ€” assists with sorting evidence, building arguments, transcribing and distinguishing speakers during hearings, and rapidly assessing the latest developments in a case.
๐Ÿ›ก๏ธ Governance & Compliance โ€” supports the implementation of GDPR, the EU AI Act, NIS2 and internal policies, including verification of legal citations against a proprietary database and the creation of audit materials.
5

๐Ÿ›ก๏ธ AML, OSINT, Fraud & Security

๐Ÿฆ AML & Sanctions Screening โ€” screens individuals, companies and properties against sanctions lists, PEP registers, insolvencies, enforcement orders, UBO records and other registries; includes 18 modules, a risk score, a status lifecycle and a structured risk report.
๐Ÿ” OSINT & Due Diligence โ€” traces connections, public digital footprints and verifiable information about a subject, converting them into a clear briefing document for commercial, security or legal investigations.
๐Ÿšจ CzechAI Shield & Fraud-Guard โ€” supports the detection of insurance and other fraud through case prioritisation, anomaly detection, network link analysis and a calculated fraud score.
๐Ÿ”’ Security Guardrails โ€” protects AI operations by detecting sensitive data, defending against prompt injection, managing roles and access permissions, logging activity and generating auditable security documentation.
6

๐Ÿฆ Banks, Finance & Insurance

๐Ÿ  Collateral Valuation โ€” model V26 prepares a market estimate, price range and comparative materials for LTV, mortgage, credit decision and collateral review.
๐Ÿ“‹ Underwriting and claims settlement โ€” sorts cases, reads documentation via OCR, supplements valuations and external context, and prepares materials for faster decision-making by the responsible officer.
๐Ÿ’ณ Loans and financing โ€” covers credit scoring, creditworthiness and mortgage calculations, and related financial workflows; the proprietary platform HypoStars at hypostars.cz offers a calculator and a partner portal.
๐Ÿ“‰ Kronos Financial Prediction โ€” uses ML models for financial and market predictions and provides analytical background, not individual investment recommendations.
๐Ÿง‘โ€๐Ÿ’ผ Banker Copilot โ€” works as an assistant directly within the CRM, searches product knowledge, calculates LTV and DSCR, and prepares materials, summaries and client correspondence.
๐Ÿ  Collateral Intelligence โ€” continuously monitors collateral values, revalues the entire portfolio and provides timely alerts on value decline, risk changes or the need for further review.
๐Ÿ“Š Insurance Knowledge Base โ€” consolidates insurance terms and conditions, exclusions, limits and valuation tables into a single searchable knowledge base with AI-powered search.
๐Ÿšจ Anti-fraud engine and claims settlement โ€” detects anomalies, prioritises suspicious cases and supports claims settlement through automatic reading, classification and verification of documents.
โ™ป๏ธ Reinsurance and catastrophe modelling โ€” prepares materials for reinsurance and models cumulative risks, exposure concentrations and the potential impact of large-scale loss events.
๐Ÿ›๏ธ ฤŒNB reporting and institution database โ€” creates data materials for regulatory reporting and maintains a structured register of banks, insurance companies and other financial institutions.
โš™๏ธ Regulated deployment โ€” we design on-premise or private cloud architecture with emphasis on data protection, auditability, access management and secure integration into the institution's processes.
7

๐Ÿ›๏ธ State ยท 16 AI solutions for public administration

๐Ÿงพ Public data analytics โ€” a set of 16 AI solutions built on official, tax, property and open data, helping institutions find connections and prepare materials more quickly.
๐Ÿ“Œ Official notice boards and opportunities โ€” automatically collects, sorts and filters notices, auctions, sales, grant opportunities and public procurement notices published by municipalities and institutions.
๐Ÿ‘ฎ Supporting analytical tools โ€” the Police AI platform and other systems accelerate work with information, transcripts, documents and relationships, while always preserving the decision-making authority of the responsible person.
๐Ÿงพ AI tax detective โ€” cross-references data from the financial administration, land registry and public registers to identify discrepancies or risk patterns that can serve as a basis for further investigation.
๐Ÿ›ฐ๏ธ Official notice board monitoring โ€” continuously monitors new postings across municipalities, regions, ministries and other institutions, and delivers relevant changes in the form of alerts.
๐Ÿ›ก๏ธ Cyber defence AI and transport infrastructure โ€” supports attack detection and forensic analysis while enabling prediction of traffic load and better planning of infrastructure investments.
โ˜๏ธ Gov Cloud โ€” multi-tenant AI infrastructure for public institutions with separated data, an agent catalogue, permission management and secure connection to state registries.
๐ŸŽ“ AI education for institutions โ€” CzechAI Academy offers practical training, AI literacy, governance, security and the use of AI in accordance with the role and responsibilities of the organisation.
8

๐Ÿ“ฃ Marketing, Facebook and Reach

๐ŸŽ›๏ธ Marketing Command Deck โ€” manages a team of AI specialists covering strategy, content creation, acquisition, analytics, distribution and marketing process automation.
๐Ÿ‘ฅ Largest Facebook Community in the Czech Republic โ€” 230,000+ Members โ€” owns the largest Czech Facebook group in the field, complemented by a network of 241 community channels, of which 92 cover investments, real estate and finance, enabling organic reach, audience activation and targeted content distribution.
๐Ÿ“ฌ Rozesรญlka.eu โ€” 37 Million Contacts โ€” enables bulk e-mail campaigns with segmentation by industry, region and entity type via a proprietary mail server and campaign tool.
๐Ÿ“ˆ Performance, Content and Distribution Network โ€” connects copywriting, SEO/GEO, paid media, social networks, e-mail and SMS lifecycle, CRO and reporting into a single measurable acquisition system.
โœ‰๏ธ Rozesรญlka.eu โ€” a proprietary platform for working with data, campaigns and AI communication, providing segmentation, bulk distribution, personalisation and response measurement.
๐Ÿ” CRM, Leads and Sales Workflow โ€” finds, enriches and qualifies leads, records them in the CRM, manages follow-ups and provides the sales team with ongoing reporting.
9

๐ŸŽ™๏ธ Voice, Image, Video and Documents

๐Ÿ—ฃ๏ธ Voice Assistants โ€” communicate in natural Czech, handle telephone and conversational scenarios, transcribe calls and distinguish individual speakers.
๐Ÿง‘โ€๐Ÿ’ผ Avatars and Video โ€” create talking digital assistants, product and training videos, and an automated video pipeline from script to final content.
๐Ÿ–ผ๏ธ Visual and 3D Creation โ€” encompasses image generation, product visuals, 3D elements, interactive maps and presentation materials for both marketing and professional use.
๐Ÿ“‘ Documents and Reports โ€” automates OCR, processing of PDFs, contracts, reports, flipbooks and data outputs, as well as the preparation of ready-made client materials.
๐Ÿ–จ๏ธ Digitรกlnรญ Tiskรกrna โ€” the tiskarna.digital platform converts digital outputs into physical products such as printed materials, custom printing, merchandising and bespoke presentation materials.
10

๐ŸŒ Applications, Portals and Integrations

๐Ÿ’ฌ CzechAI Chat, Agent MAX, Aurum and darkstar โ€” agentic chatbots provide access to models, tools, research, documents and specialised assistants within a single user environment.
๐Ÿ˜๏ธ EstateX and Vertical Portals โ€” ready-made platforms for real estate, finance, education, marketing and client self-service are complemented by the Digital Marketplace at trziste.digital.
๐Ÿ”Œ Custom Integrations โ€” we connect APIs, MCP servers, n8n, e-mail, CRM, databases, spreadsheets, documents and other client systems into a single functional workflow.
๐Ÿ› ๏ธ Turnkey Websites and SaaS โ€” we deliver the complete product from the first working version through to live operation, including frontend, backend, payments, administration, monitoring and ongoing development.
11

โš™๏ธ Infrastructure, Operations and DevOps

๐Ÿณ Production Operations โ€” we manage servers, GPUs, Docker, PM2, databases, model services, monitoring and operational automations across 1,500 services, 273 containers and 6 servers.
โ˜๏ธ Private Cloud and On-Premise โ€” we design infrastructure according to data sensitivity, regulatory requirements, required performance and the method of integration into the client's environment.
๐Ÿ“ก Reliability and oversight โ€” we ensure availability monitoring, backups, centralised logging, alerting and operational procedures for rapid incident resolution.
๐Ÿš€ Fast delivery โ€” we can prepare the first functional MVP for a specific problem within weeks and continue expanding it based on measured results and actual user needs.
12

๐ŸŽ“ Education, strategy and transformation

๐Ÿซ CzechAI Academy โ€” academy.czechai.io is a proprietary certification and education programme for individuals, teams, businesses and public institutions, focused on the practical and safe use of AI.
๐Ÿงญ AI strategy โ€” helps select priorities, use cases, expected value, risks, data readiness and AI governance approach so that the investment leads to a measurable outcome.
๐Ÿ‘ฅ Team adoption โ€” through practical workshops, internal assistants, guides and support, we set up ways of working so that employees genuinely embrace and use AI on a daily basis.
๐Ÿ“Š AI audit and roadmap โ€” assesses existing solutions, data, processes and costs, and translates findings into concrete priorities, measurable goals and a realistic deployment plan.
๐Ÿ” AI X-Ray โ€” independently examines third-party AI solutions, data and infrastructure, and precisely shows what works, what is missing, where risks arise and what is being paid for unnecessarily.
๐ŸŽ“ Certifications L1โ€“L5 โ€” covers the journey from AI literacy through practitioner and automation specialist to governance and domain specialisation; the corporate package includes the entire team and an internal AI policy.
๐ŸŽค Simulators and training โ€” a sales simulator works with model customer types, and a training assistant helps financial advisors safely practise real-life situations.
๐Ÿซ Educational AI for schools โ€” a teacher assistant prepares lessons, tests and assessment materials, while the educational AI communicates by voice and works with cited sources.
13

๐Ÿงฐ Sole traders and small businesses

๐Ÿ“จ Contract generator โ€” continuously searches for public and commercial tenders and opportunities matching the trade, region and capacity of a specific entrepreneur.
๐Ÿ’ต Price spy โ€” monitors competitor prices in the area, compares offers and helps set pricing that is both competitive and profitable.
๐Ÿงพ Autonomous invoicing agent โ€” automatically issues invoices, monitors due dates, flags unpaid documents and prepares or sends payment reminders.
โš–๏ธ Pocket lawyer โ€” assists with contracts, complaints, receivables and basic legal situations that sole traders encounter in their day-to-day business.
โ˜Ž๏ธ Voice receptionist โ€” answers calls, responds to routine enquiries, books appointments and takes messages, so the business never misses a customer even outside working hours.
๐Ÿ’ถ Grant radar โ€” monitors newly announced grants and subsidies, filters them by company profile and alerts users to suitable opportunities and approaching deadlines.
โญ Reputation manager โ€” tracks reviews and ratings across platforms, flags new responses and helps systematically build brand credibility.
14

๐Ÿข Corporate sector

๐Ÿ”Ž Company Screening Platform โ€” verifies business partners, suppliers and counterparties against registries and open data, and creates a unified risk profile for each entity.
๐Ÿ”— Supply Chain Risk Monitor โ€” continuously monitors insolvencies, ownership changes, sanctions, and other risk events in the supply chain, providing timely alerts about at-risk suppliers.
๐ŸŽฏ M&A Target Finder โ€” searches for suitable acquisition targets based on industry, location, financial indicators, ownership structure, and the investor's strategic criteria.
๐ŸŒฑ ESG Compliance Platform โ€” supports reporting under CSRD and ESRS, collects and consolidates ESG data, and creates traceable documentation for internal and external reporting.
๐Ÿ›’ Procurement Optimizer โ€” analyses purchasing activity, compares prices, evaluates suppliers, and identifies opportunities for savings, order consolidation, and improved contract terms.
๐Ÿ‘ค HR Screening & Background Check โ€” vets candidates exclusively using legitimately accessible public and registry sources, and prepares structured documentation to support responsible personnel decisions.
๐Ÿ“Š Business Intelligence Hub a Knowledge Management โ€” connects company data with an internal knowledge base, enabling employees to search, analyse, and utilise information with the help of AI.
15

๐Ÿ›’ E-commerce

๐Ÿ’ฐ Competitive Pricing Engine โ€” continuously compares prices across e-shops and provides the data needed for automatic or managed dynamic pricing.
๐Ÿ“‰ Churn Prediction and Customer Segmentation โ€” predicts customer churn, creates RFM segments, and enables more precise targeting of both retention and acquisition campaigns.
๐ŸŽ Product Recommendation AI โ€” recommends relevant products to each customer based on their behaviour, purchase history, item similarity, and current context.
๐Ÿšจ Fraud Detection โ€” identifies suspicious orders, unusual payment patterns, and potential misuse of accounts, coupons, or returns processes.
๐Ÿ“ฆ Inventory Optimizer a Demand Forecasting โ€” forecasts demand, recommends optimal stock levels, and reduces stockouts, surpluses, and unnecessarily tied-up capital.
๐Ÿšš Shipping Cost Optimizer โ€” compares the price, speed, and quality of delivery across ฤŒeskรก poลกta, PPL, DPD, Zรกsilkovna, and other carriers, selecting the most advantageous option.
๐Ÿ›๏ธ Abandoned Cart Recovery, Review & Reputation Manager a Marketing Attribution โ€” brings customers back to abandoned carts, manages reviews, and shows which marketing channels actually drive orders.
16

๐Ÿšš Logistika a doprava

๐Ÿ—บ๏ธ Route Optimization Engine โ€” optimises delivery and collection routes using TSP/VRP algorithms on the precise RรšIAN address system, taking into account time, capacity, and stop sequencing.
๐Ÿš› Fleet Management a Carrier Performance โ€” manages the vehicle fleet, monitors vehicle utilisation, and evaluates carriers by price, reliability, speed, and complaints.
๐Ÿญ Warehouse Picking Optimizer a Cross-Dock Optimizer โ€” reduces movement within the warehouse, optimises picking sequences, and plans the transfer of goods without unnecessary storage.
๐Ÿ“ Load Planning Optimizer โ€” uses 3D bin packing to plan cargo placement so as to maximise vehicle capacity utilisation while respecting weight and operational constraints.
โฑ๏ธ Delivery Time Predictor โ€” estimates the actual delivery time based on route, traffic, carrier history, warehouse load, and other current conditions.
โ„๏ธ Cold Chain Monitor โ€” monitors the temperature and progress of shipments containing sensitive goods, and alerts to any cold chain breach or risk of consignment spoilage.
โ†ฉ๏ธ Returns Management โ€” automates the recording, sorting, approval and evaluation of returns, and helps identify their most common causes as well as potential misuse.
17

๐Ÿญ Manufacturing

๐Ÿ“ˆ Production Planning Optimizer โ€” plans production according to orders, capacities, materials and deadlines, and monitors OEE to identify downtime, performance losses and bottlenecks.
๐Ÿ” Quality Control AI โ€” detects defects using computer vision and integrates visual inspection with statistical process control (SPC).
๐Ÿ”ง Predictive Maintenance โ€” analyses data from IoT sensors and operational history, and predicts failures before they cause downtime or a costly breakdown.
๐ŸŒ Digital Twin Simulator โ€” creates a digital twin of the production line for safe testing of changes, capacities, scenarios and impacts before any intervention in the real operation.
๐Ÿท๏ธ Traceability System โ€” ensures track & trace of materials, products and batches, and provides precise traceability during quality inspections or recall actions.
โšก Energy Consumption Optimizer a Yield Optimization โ€” optimises energy consumption and production yield based on machines, shifts, products and current operating conditions.
๐Ÿ‘ท Workforce Optimization a Continuous Improvement Tracker โ€” schedules shifts according to required skills and performance, and records improvement proposals in Kaizen and Lean mode.
18

๐Ÿ’ผ Professional Services

๐Ÿ’น Project Profitability Analyzer โ€” tracks actual costs, time, billing and margin for individual projects, and provides early warnings for projects that are becoming unprofitable.
๐Ÿ“ Proposal Generator AI a Document Automation โ€” creates business proposals, contractual annexes and recurring documents from verified templates and available client data.
๐Ÿ—“๏ธ Resource Capacity Planner โ€” plans team capacity according to projects, roles, deadlines and workload, and helps prevent both overload and underutilised staff.
๐Ÿงพ Invoice Automation โ€” connects completed work with billing, automatically prepares documents and monitors their dispatch, due dates and payment.
๐Ÿ“š Knowledge Base AI โ€” converts company documents, expertise and methodologies into a central knowledge base with fast AI-powered search and traceable sources.
๐Ÿ˜Š Client Satisfaction Tracker โ€” continuously measures client satisfaction, NPS and feedback, and flags relationships that require a prompt response.
๐Ÿค Client Relationship Manager a Business Development CRM โ€” manages relationships, contacts, communication history, business opportunities and the entire pipeline from first outreach to closed deal.
19

๐Ÿฅ Healthcare

๐Ÿ—“๏ธ Patient Appointment Optimizer โ€” optimises patient scheduling, surgery utilisation and waiting times, and predicts the risk of a patient failing to attend their appointment.
๐ŸŽ™๏ธ Medical Transcription AI โ€” transcribes medical dictations and conversations into structured documentation and supports the assignment of ICD-10 codes for subsequent expert review.
๐Ÿฉบ Clinical Decision Support โ€” searches specialist sources using RAG and prepares relevant supporting materials; it never replaces the clinical judgement of a qualified physician.
๐Ÿ“„ Insurance Claim Automation โ€” automates the processing of health insurance claims and documentation checks using ICD-10 and CPT classifications.
โš ๏ธ Patient Risk Stratification and Medication Adherence โ€” segments patients by risk level and helps monitor treatment adherence, enabling timely focus of care on the most vulnerable groups.
๐Ÿ‘ฉโ€โš•๏ธ Staff Scheduling Optimizer โ€” schedules healthcare staff shifts based on specialisation, availability, regulatory constraints and anticipated workload.
๐Ÿ“Š Healthcare Supply Chain and Revenue Cycle Analytics โ€” tracks supply, consumption, billing and operational economics, identifying opportunities for savings without compromising care availability.
03

๐Ÿš€ II. TECHNOLOGY, PRODUCT AND DEVELOPMENT AREAS

This section preserves individual technological, product and development objectives as standalone entries. Each line represents a specific capability, development direction or operational area; items are not merged into general summaries.
T1

๐Ÿ–ฅ๏ธ GPU, Servers and Model Infrastructure

๐Ÿ–ฅ๏ธ Expansion of proprietary GPU and server infrastructure in the Czech Republic โ€” scaling computational capacity for AI inference, fine-tuning, machine learning, data processing, content generation and rendering.
๐ŸŽ›๏ธ Acquisition and operation of high-performance GPUs โ€” building proprietary hardware foundations for language, image, voice, analytics and multimodal models.
โšก Scalable computational capacity โ€” delivering processing power to internal projects and external customers in line with current demand.
๐Ÿงฎ GPU compute rental โ€” commercial provision of secured computational power for inference, training, fine-tuning, processing and rendering.
๐Ÿ”Œ Model APIs โ€” exposing proprietary, open-source and external models through a unified, secured interface.
๐Ÿšฆ Central model gateway โ€” managing proprietary, open-source and external language models from a single operational point.
๐Ÿงญ Intelligent task routing โ€” automatic selection of the most suitable model based on cost, speed, data sensitivity, domain and task complexity.
โš–๏ธ Load balancing across models and GPU servers โ€” distributing traffic according to available capacity, latency and current load.
๐Ÿ›Ÿ Automatic failover โ€” transferring a task to another model or server in the event of failure, unavailability or breach of an operational limit.
๐Ÿ‡จ๐Ÿ‡ฟ Czech sovereign AI platform โ€” proprietary and open-source models, operated in the Czech Republic with independence from any single foreign provider.
๐Ÿข Private cloud โ€” a dedicated environment designed around data sensitivity, required performance and the specific organisation's policies.
๐Ÿ  On-premise deployment โ€” operation directly within the customer's infrastructure, where sensitive data need never leave the organisation.
๐Ÿ”‹ Energy optimisation of GPU infrastructure โ€” managing power consumption, load, cooling and available energy utilisation across compute nodes.
๐Ÿงช GPU and HPC testing environment โ€” research, development, benchmarking, pilot deployments and model validation on high-performance infrastructure.
T2

๐Ÿง  Models, RAG and Knowledge Systems

๐Ÿง  Secure operation of proprietary language models โ€” controlled model deployment under full in-house governance, with the ability to customise performance, data handling and security policies.
๐ŸŒ Secure operation of open language models โ€” deploying open-source models without dependency on a single commercial API.
๐Ÿงช Model fine-tuning โ€” further training of models on prepared data for a specific domain, organisation, or work task.
๐Ÿ‡จ๐Ÿ‡ฟ Model specialisation for Czech โ€” improving handling of the Czech language, declension, meaning, style, and local context.
โš–๏ธ Legal and professional terminology โ€” adapting models to Czech legislation, case law, and the specialist terminology of individual sectors.
๐Ÿ“„ Adaptation to corporate documentation โ€” tailoring responses to the customer's internal documents, methodologies, products, and processes.
๐Ÿ› ๏ธ Tool-calling and working with tools โ€” developing models capable of securely calling APIs, applications, registries, databases, and automation workflows.
๐Ÿ“š RAG systems โ€” precise retrieval and question-answering over extensive documents, databases, and the organisation's current knowledge.
๐Ÿ”Ž Hybrid search โ€” combining full-text, metadata, relational filters, vector similarity, and domain-specific scoring.
๐Ÿ•ธ๏ธ Knowledge graphs โ€” linking persons, companies, documents, events, properties, legal relationships, and other entities.
โœ… Verifiable answer sources โ€” automatic citations, traceability of information used, and the option of subsequent expert review.
๐Ÿงฉ Czech knowledge layer cz-rag โ€” a dedicated layer for working with Czech documents, registries, legislation, and domain-specific data.
๐Ÿ—„๏ธ Vector databases โ€” storing and retrieving embeddings over case law, legislation, documents, and other knowledge collections.
๐Ÿงฐ Data and knowledge technologies โ€” PostgreSQL, Qdrant, Neo4j, Weaviate, ClickHouse, Redis, Kafka, MongoDB, and other database, streaming, and analytics layers.
T3

๐Ÿ—ƒ๏ธ Data, registries, OCR, geodata and analytics

๐Ÿ•ท๏ธ Automated data collection โ€” gathering information from websites, documents, public sources, registries, and client systems.
๐Ÿงน Data cleansing and normalisation โ€” converting diverse inputs into a unified, usable, and auditable structure.
๐Ÿงฌ Deduplication and matching โ€” removing duplicates and linking records belonging to the same person, company, document, or object.
๐Ÿ•ถ๏ธ Anonymisation of sensitive data โ€” protecting personal and confidential information during data processing, analysis, and transfer.
๐Ÿท๏ธ Data classification and enrichment โ€” adding categories, relationships, context, scores, and other derived attributes.
๐Ÿ”„ Continuous updates and archiving โ€” preserving change history and regularly refreshing data from live sources.
โšก Real-time data pipelines โ€” continuous monitoring, validation, deduplication, archiving and updating of millions of records.
๐Ÿ›๏ธ ARES โ€” company and economic data usable for due diligence, contact enrichment, analytics and business processes.
๐Ÿ—บ๏ธ ฤŒรšZK and cadastral data โ€” land parcels, title deeds, ownership relationships and other resources for real estate and asset analysis.
๐Ÿ“ RรšIAN โ€” precise address and territorial system for geocoding, logistics, mapping and spatial analysis.
โš ๏ธ ISIR, auctions and official notice boards โ€” insolvencies, forced sales, published documents and other events relevant to opportunity and risk monitoring.
โš–๏ธ e-Sbรญrka and judicial databases โ€” regulations, statutory provisions, case law and verifiable legal sources.
๐Ÿ›ก๏ธ Sanctions and reference lists โ€” sources for AML screening, PEP checks, relationship verification and risk management.
๐Ÿ”ค OCR in more than 90 languages โ€” conversion of scanned documents, photographs and forms into machine-readable text.
๐Ÿ“„ Document parsers โ€” extraction of data, tables, attachments, metadata and the logical structure of documents.
๐ŸŽ™๏ธ Transcription and diarisation of recordings โ€” conversion of audio to text and identification of individual speakers.
๐Ÿ”€ Statement comparison โ€” identification of matches, differences, contradictions and connections across multiple transcripts or documents.
๐Ÿ“‘ Automated document outputs โ€” generation of PDFs, DOCX, XLSX, reports, proposals, contractual annexes and client materials.
๐Ÿ—บ๏ธ Geocoding and GIS โ€” conversion of addresses and parcels into spatial data, layer management and location intelligence.
๐Ÿ”ฅ Heatmaps, risk layers and 3D visualisations โ€” clear visualisation of prices, events, risks, infrastructure and territorial relationships.
๐ŸŒ Global monitoring worldmonitor โ€” continuous geopolitical, economic, security and event-based monitoring of public and specialist sources.
T4

๐Ÿค– Agents, MCP, integrations and process automation

๐Ÿค– Autonomous AI agents โ€” use tools, execute real-world tasks and trigger follow-on processes.
๐Ÿ‘จโ€๐Ÿ’ป Specialised domain agents โ€” roles for development, law, finance, real estate, marketing, security, research, design, data and operations.
๐Ÿค Domain AI swarms โ€” coordinated teams of specialists assembled according to the type of task and desired outcome.
๐Ÿง  Long-term agent memory โ€” retention of relevant context, decisions and work-in-progress outputs.
๐Ÿ” Handover of work between agents โ€” automatic continuation of a task by a different role without loss of essential context.
๐ŸŽ›๏ธ Central orchestration โ€” recognising the assignment, dividing the work, selecting specialists, coordinating the process and reviewing the result.
๐Ÿ”Œ MCP server development โ€” standardised tool bridges for secure AI interaction with real-world systems.
๐ŸŒ API interface development โ€” connecting models and agents to products, registries, client applications and external services.
๐Ÿงฉ Connector development โ€” linking AI with email, CRM, ERP, accounting, databases, documents, spreadsheets and web applications.
โš™๏ธ Automation of real business processes โ€” the journey from receiving a request, through finding and verifying information, to a decision-support document, system update or dispatched output.
๐Ÿ—‚๏ธ Central working-layer catalogue โ€” a registry of agents, skills, commands, tools, MCP servers and pre-built teams.
๐Ÿ’ฌ Enterprise AI assistants and copilots โ€” CzechAI Chat, Agent MAX, Aurum AI and darkstar with access to models, internal knowledge, databases, tools and live data sources.
๐Ÿ”— n8n, CRM and downstream workflow integration โ€” triggering repeatable processes across applications without manual data re-entry.
T5

๐Ÿ›ก๏ธ Security, audit, governance and regulated operations

๐Ÿ” Data and communication encryption โ€” protecting information in transit, at rest and during processing.
๐Ÿชช Identity, role and permission management โ€” ensuring users and agents can only access permitted data, tools and actions.
๐Ÿ•ถ๏ธ Sensitive data protection โ€” detecting personal, confidential and regulated information before further use.
๐Ÿงฑ Customer isolation โ€” segregation of data, context, model inputs and operating environments.
๐Ÿšซ Prompt injection protection โ€” detecting and mitigating attempts to alter system rules or exploit connected tools.
๐Ÿ›ก๏ธ Security guardrails โ€” technical and procedural boundaries for working with code, data, documents and critical systems.
๐Ÿงพ Audit trail โ€” traceable records of inputs, outputs, sources used, tools invoked and actions taken.
๐Ÿ“ก Logging and monitoring โ€” tracking availability, performance, errors, changes and security events.
โœ… Model testing and validation โ€” verifying the correctness, stability, safety and suitability of outputs for specific use cases.
๐Ÿ” Hallucination and unsupported output control โ€” verifying claims against sources, databases and system rules.
โš–๏ธ AI governance and regulatory compliance โ€” EU AI Act, GDPR, NIS2, ISO/IEC 42001 and related internal obligations.
๐Ÿ“‹ AI system registry and risk management โ€” a catalogue of solutions in use, responsibilities, risks, controls, and lifecycle management.
๐Ÿ“ Corporate AI Policy โ€” rules for the safe, responsible, and auditable use of AI within the organisation.
๐Ÿ‘ค Human in the loop โ€” AI prepares and validates materials, but the final decision remains with an authorised person.
T6

๐Ÿญ Products, platforms and industry systems

๐Ÿ  Reality Engine โ€” continuous collection, consolidation, and updating of real estate listings, auctions, insolvencies, enforcement proceedings, state offerings, official notice boards, and cadastral data.
๐Ÿ“ˆ Price Map V26 and proprietary valuation models โ€” multi-level AI valuation of apartments, houses, land, recreational properties, and commercial real estate.
๐Ÿ“Š Comparative real estate analyses โ€” price ranges, comparable listings, reliability ratings, and explanations of the underlying data sources used.
๐Ÿ’Ž Undervalued property search โ€” offer scoring, market comparison, and automatic alerts for investment opportunities.
๐Ÿ‘ค Direct owners and off-market listings โ€” sourcing opportunities outside standard real estate distribution channels.
โณ Long-listed and recurring listings โ€” tracking advertising history, price changes, and signals of motivated selling.
๐Ÿ”จ Auctions, insolvencies, enforcement proceedings, and state sales โ€” monitoring of forced and non-standard sales situations.
๐Ÿ—บ๏ธ Territorial development forecasting โ€” evaluation of potential changes in land use based on zoning plans, รšAP, ZรšR, รšSES, and other sources.
๐Ÿ˜๏ธ EstateX โ€” a platform for real estate agencies, brokers, investors, and other professionals.
๐Ÿ” RealityScan โ€” verification, collection, and analytical processing of real estate information.
๐Ÿค– Reality Copilot โ€” an AI workspace for property search, valuation, due diligence, documents, clients, and business decision-making.
๐Ÿ“ฑ Essty โ€” a mobile real estate and personal AI assistant with voice control, valuation, client management, and live data.
๐Ÿ“ก Lead Hub โ€” delivery of direct owners, undervalued listings, auctions, insolvencies, and other leads into a spreadsheet, API, or CRM.
๐Ÿ• Property Watchdog โ€” continuous monitoring of properties, price changes, listing status, and new risk or commercial signals.
๐ŸŽฏ Opportunity Engine โ€” automated discovery, evaluation, and prioritisation of investment opportunities.
๐Ÿข HOA and property manager application โ€” resident communication, billing, fault reporting, reserve fund management, and tradesperson dispatch.
๐Ÿฆ HypoStars โ€” mortgages, loans, bank comparison, affordability calculation, scoring, and partner collaboration.
๐Ÿงฎ Banking AI tools โ€” collateral valuation, LTV and DSCR calculations, portfolio monitoring, and preparation of credit documentation.
๐Ÿ“‹ Insurance underwriting โ€” case sorting, reading documentation, supplementing external context and preparing materials for decision-making.
๐Ÿšจ Claims settlement and anti-fraud โ€” automated document reading, anomaly detection, prioritisation of suspicious cases and network connections.
โ™ป๏ธ Reinsurance and catastrophe modelling โ€” management of accumulated risks, exposure concentrations and regulatory documentation.
๐Ÿ’น Financial analytics and predictive models โ€” scoring, corporate analysis, market and financial forecasting and automated reporting.
โš–๏ธ Legal AI โ€” Czech legislation, case law, legal research, contracts, redlines, submissions, dispute timelines and citation verification.
๐ŸŽ™๏ธ Transcription of court hearings and interrogations โ€” transcription, diarisation, comparison of testimonies and preparation of analytical materials.
๐Ÿ‘ฎ Police AI โ€” recording of interrogations, identifying contradictions, working with information and supporting analysis without replacing the responsible person.
๐Ÿงพ AI tax detective โ€” cross-referencing tax, property, corporate, cadastral and other public data.
๐Ÿ›๏ธ Systems for public administration โ€” official notice boards, file management, forms, public procurement, property data and control processes.
โ˜๏ธ Gov Cloud โ€” secured multi-tenant infrastructure with an agent catalogue and integration with public administration systems and registries.
๐Ÿ›ก๏ธ AML platform โ€” sanctions, PEP, insolvency, enforcement, ownership structures, UBO, financial data, risk score and structured reports.
๐Ÿ” OSINT & Due Diligence โ€” public registries, digital footprint, entity connections, domains, contacts, documents and international databases.
๐Ÿšจ Anti-fraud and security systems โ€” financial anomalies, suspicious transactions, service abuse, insurance fraud and risk-related connections.
๐Ÿ–ฅ๏ธ Cyber defence AI โ€” detection of attacks, security events and support for forensic analysis.
๐Ÿ“ฃ Marketing Command Deck โ€” orchestration of specialists for strategy, content, acquisition, advertising, analytics and automation.
๐ŸŽฏ Lead and CRM systems โ€” search, enrichment, segmentation, scoring, qualification and automatic transfer of contacts to CRM.
๐Ÿ“ฌ Rozesรญlka.eu โ€” email, SMS and multi-channel campaigns, communication automation and response measurement.
๐Ÿ›’ E-commerce tools โ€” dynamic pricing, product recommendations, inventory, demand forecasting, fraud detection, abandoned carts, reviews and attribution.
๐Ÿšš Logistics systems โ€” route optimisation, load planning, vehicle fleets, warehouses and delivery forecasting.
๐Ÿญ Manufacturing systems โ€” capacity planning, quality control, computer vision, predictive maintenance, digital twins and traceability.
โšก Energy systems โ€” optimisation of energy consumption in buildings, manufacturing facilities, data centres and GPU infrastructure.
๐Ÿฅ Healthcare applications โ€” documentation transcription, capacity planning, working with specialist sources and support for qualified decision-making.
๐Ÿงฐ Tools for sole traders and small businesses โ€” invoicing, receivables, orders, contracts, grants, reputation management and virtual reception.
๐Ÿข Corporate Systems โ€” supplier screening, supply-chain risk, ESG, procurement, HR and acquisition target search.
๐Ÿ“Š Business Intelligence and Knowledge Management โ€” corporate data layers, knowledge bases, AI search and management dashboards.
T7

๐ŸŽ™๏ธ Voice, Multimedia, Documents and Application Development

๐Ÿ—ฃ๏ธ Czech Voice Models โ€” voice synthesis and speech recognition tailored to the Czech language.
โ˜Ž๏ธ Phone Agents and Virtual Receptionists โ€” call handling, responses, appointment scheduling and message forwarding.
๐ŸŽง Customer Support โ€” conversational client service, request triage and escalation of complex cases.
๐Ÿ—ฃ๏ธ Voice Control of Applications โ€” controlling features and workflows using natural speech.
๐Ÿ–ผ๏ธ Image and Product Visualisation Generation โ€” marketing, presentation, product and professional image outputs.
๐ŸŽฌ Video, Avatars and Educational Content โ€” talking assistants, automated video pipelines and multimedia learning.
๐ŸŽต Music and Audio Outputs โ€” generation and processing of music, voice and other audio materials.
๐ŸงŠ 3D Elements and Interactive Content โ€” visualisations, models, maps, presentations and other multimodal outputs.
๐Ÿ“š Personalised Books and Materials โ€” automated preparation of content tailored to a specific user or target audience.
๐ŸŒ Turnkey Web Applications โ€” frontend, backend, authentication, administration, payments, monitoring and long-term operation.
๐Ÿ“ฑ Mobile Applications โ€” native or cross-platform products integrated with AI, data, voice and client systems.
โ˜๏ธ SaaS Products and Client Portals โ€” scalable applications, self-service, dashboards and industry platforms.
๐Ÿท๏ธ White-Label Solutions โ€” ready-made applications, agents, data platforms and industry systems operated under the customer's brand.
๐Ÿ›๏ธ Digital Marketplace โ€” offering digital products, applications, services, data outputs and automated solutions.
๐Ÿ–จ๏ธ Digital Print Shop and AI Design Studio โ€” personalised printing, books, promotional materials, custom printing, merch and print-on-demand production.
T8

๐ŸŽ“ Research, Education, Audit and Transformation

๐Ÿ” AI X-Ray โ€” independent audit of third-party AI solutions, models, data, security, infrastructure and regulatory readiness.
๐Ÿ›ก๏ธ Security and Technology Audits โ€” verification of architecture, configuration, dependencies, operations, models and data flows.
๐Ÿงญ AI strategy โ€” selection of priorities, use cases, expected value, risks, data readiness and governance approach.
๐Ÿ—บ๏ธ Transformation Roadmaps โ€” concrete steps, measurable goals, responsibilities and an AI implementation timeline.
๐ŸŽ“ CzechAI Academy โ€” an educational and certification programme for individuals, teams, businesses and public institutions.
๐Ÿญ Strojovna โ€” hands-on work with tools, models, automation, RAG, agents and real-world industry tasks.
๐Ÿชœ Certifications L1โ€“L5 โ€” AI Literacy, practical AI use, automation, governance, security and industry specialisations.
๐ŸŽค Simulators and training applications โ€” tools for sales professionals, financial advisors, employees, management and public institutions.
๐Ÿงช Research, development and pilot deployment โ€” experiments, prototypes, GPU and HPC testing, and the transfer of results into production.
๐Ÿค Collaboration with universities and research institutions โ€” joint development, testing, expert validation and knowledge transfer.
๐Ÿ“‘ Technical documentation and product certification โ€” preparation of materials for operation, audit, handover, verification and formal certification.
๐Ÿ›ก๏ธ Protection of source code, trademarks and know-how โ€” legal, technical and operational protection of created value.
T9

๐Ÿ’ผ Commercialisation, distribution and operational scaling

๐Ÿ“ฆ Licences โ€” commercial provision of software, models, data layers and specialised modules.
โ˜๏ธ SaaS โ€” delivery of ready-made services and applications on a subscription basis.
๐Ÿ”Œ API services โ€” paid access to models, data, computing, verifications, scoring and automated functions.
๐Ÿ  Private cloud and on-premise installation โ€” standalone deployment for businesses, banks, insurance companies, the state and regulated organisations.
โšก Provision of AI computing power โ€” inference, fine-tuning, computation, content generation and rendering for external customers.
๐Ÿ“Š Commercial data outputs โ€” analytical reports, monitoring, alerts, scoring and industry data services.
๐Ÿค Partner and distribution network โ€” collaboration with banks, insurance companies, real estate agencies, advisors, integrators and public institutions.
๐Ÿ‡ช๐Ÿ‡บ European partnerships and consortia โ€” internationalisation, joint projects, research and commercial development within the European space.
๐Ÿ‘ฅ Team expansion โ€” growth of development, security, research, operations and business capabilities.
๐Ÿณ Container and DevOps infrastructure โ€” secure deployment, updates and scaling of applications, data services and models.
๐Ÿ“ก Continuous availability monitoring โ€” oversight of services, performance, errors, capacity and dependencies.
โ™ป๏ธ Automatic service recovery โ€” restart, failover and restoration of operations without unnecessary manual intervention.
๐Ÿ’พ Backup and recovery โ€” protection of databases, configurations, documents, models and operational data.
๐Ÿšจ Alerting and incident management โ€” rapid problem detection, escalation and documented service restoration.
๐ŸŒ Unified European CzechAI ecosystem โ€” integration of Czech AI models, agents, data, applications, infrastructure, education and commercial services.
04

๐Ÿงฌ III. CLAUDE ARSENAL โ€” AGENTS, SWARMS AND TOOLS

The Arsenal is intentionally separated from customer-facing services in this master catalogue. It describes the internal working layer: who performs the work, how teams are composed, what tools they use and what operational rules they follow.
A0

๐Ÿงฌ Claude Stack ยท Arsenal

๐Ÿง  Claude orchestrator โ€” the central control brain recognises the assignment, breaks it down into steps, selects the appropriate specialists and assembles a team precisely for the given task.
๐Ÿง‘โ€๐Ÿ’ป Specialists for every domain โ€” ready-made agents cover development, data, law, finance, real estate, marketing, research, design, security and the day-to-day operations of the organisation.
๐Ÿ”ง Control of connected tools โ€” based on granted permissions, the Arsenal works with registries, documents, code, applications, databases, APIs and automation workflows, so it does not merely provide answers but actually performs the work.
๐Ÿ•ธ๏ธ Swarms, memory and RAG โ€” agent teams pass context to one another, share a knowledge layer and can be switched from Claude to a proprietary LLM as needed without changing the working process.
A1

๐Ÿงญ How the two catalogues complement each other

๐ŸŒ Complete solutions catalogue โ€” describes what CzechAI offers customers: products, applications, industry solutions, infrastructure, data and services.
๐Ÿงฌ This Arsenal catalogue โ€” describes the agents, teams, skills, tool integrations and control mechanisms from which the working layer is assembled.
A2

๐Ÿ“Š Arsenal by the numbers

๐Ÿค– 749 agent records โ€” a complete inventory of agent definitions available at the station; these are not 749 unique personas, but all registered records including duplicates and variants.
๐Ÿง  687 unique agents โ€” the number of distinct agents after removing duplicates from the total of 749 registered records; measured on 7. 8. 2026.
๐Ÿ—‚๏ธ 23 professional categories โ€” agents are classified by domain and type of work to enable rapid selection of the right specialist or assembly of an entire team.
๐Ÿงฉ 26 described swarms โ€” the catalogue directly details 9 permanent domain teams and 17 configurations designed for specific types of tasks.
๐Ÿ”Œ 46 MCP servers in the detailed section โ€” tool bridges enabling agents to work with databases, browsers, e-mail, registries, servers and proprietary services.
๐Ÿ› ๏ธ 1 009 skills, of which 997 are unique โ€” an extensive skills library loaded only when needed for a specific task; 1 003 items have their own file SKILL.md with a description.
โŒจ๏ธ 77 slash commands โ€” ready-made workflows triggered directly in the prompt using a slash, replacing the repeated manual entry of more complex instructions.
๐Ÿ”Œ 26 plugins โ€” extend the system with capabilities covering documents, development, data, design, finance, law, marketing, sales, GitHub, Stripe, Vercel and other areas.
๐Ÿช 3 automatic hooks โ€” czechai-guardrail.js, gitnexus a herdr-agent-state.ps1 monitor changes, dependencies, and the state of the agent process.
๐Ÿ“ 1 global rule set โ€” aurakit-security.md establishes safety guardrails that apply across all sessions regardless of the currently open project.
A3

๐Ÿค– Agents by profession

๐Ÿ’ป Developers and programmers โ€” 107 โ€” build frontend, backend, API, mobile, and 3D applications; work with specific languages and frameworks; and handle architecture, refactoring, and custom code implementation.
๐Ÿ—๏ธ DevOps and infrastructure โ€” 55 โ€” manage Linux, servers, Docker, Kubernetes, CI/CD, networks, monitoring, incidents, scaling, performance, and secure production operations.
๐Ÿ—„๏ธ Databases and data engineering โ€” 20 โ€” design schemas, SQL queries, data pipelines, migrations, indexes, storage, and database optimisation for reliable and fast data access.
๐Ÿ“Š Data science and AI/ML โ€” 18 โ€” perform analyses, statistical modelling, machine learning, predictions, experiments, model evaluation, and data preparation for AI.
๐ŸŽจ UI/UX designers โ€” 17 โ€” design interfaces, user journeys, design systems, visual style, accessibility, prototypes, and the usability of digital products.
โœ๏ธ Copywriters and content โ€” 56 โ€” write websites, articles, emails, ads, scripts, documentation, and specialist content, and refine tone, structure, readability, and persuasiveness.
๐Ÿ“ˆ Marketers and growth โ€” 31 โ€” handle positioning, campaigns, SEO/GEO, paid advertising, analytics, CRO, lifecycle, social media, automation, and measurable growth.
๐Ÿค Sales and business development โ€” 17 โ€” handle prospecting, lead qualification, proposals, argumentation, follow-ups, sales processes, CRM, and support for closing deals.
๐ŸŽง Customer support โ€” 6 โ€” respond to customers, triage requests, prepare knowledge-base answers, escalate issues, and help manage support quality.
๐Ÿ’ฐ Finance and accounting โ€” 20 โ€” support financial analysis, budgeting, reporting, valuation, investment research, accounting workflows, cash flow, and economic risk management.
โš–๏ธ Legal and compliance โ€” 34 โ€” analyse contracts, legal documents, GDPR, regulatory obligations, internal policies, risks, and process alignment with applicable regulations.
๐Ÿ•ต๏ธ Security and OSINT โ€” 28 โ€” conduct security audits, threat intelligence, penetration testing, background checks on individuals and companies, evidence gathering, and evaluation of public digital footprints.
๐Ÿ” Research and desk research โ€” 45 โ€” gather information from multiple sources, verify facts, work with citations, track competitors and trends, and distil findings into actionable syntheses.
๐Ÿ“‹ Project and product โ€” 14 โ€” manage briefs, backlogs, priorities, roadmaps, specifications, stakeholders, risks, and handoffs between product, development, and operations.
๐Ÿ˜๏ธ Architects and review โ€” 53 โ€” design solution structures and independently review code, security, performance, consistency, scalability, and the correctness of the resulting design.
๐Ÿงช QA and testing โ€” 16 โ€” prepare and execute automated and manual tests, identify regression bugs, verify system behaviour, and check quality prior to deployment.
๐Ÿ  Real estate and valuation โ€” 6 โ€” search for properties, assess price, investment potential, comparable listings, legal status, and other supporting materials for real estate decision-making.
๐ŸŽฌ Video, audio and creative โ€” 17 โ€” prepare scripts, visuals, editing, animation, sound, voice, 3D elements and other creative outputs for marketing, education and product communication.
๐Ÿง‘โ€๐Ÿซ HR and education โ€” 5 โ€” support recruitment, people management, training, creation of study materials, skills development and practical AI adoption within teams and schools.
๐ŸŒ Web, CMS and integrations โ€” 37 โ€” build and manage websites, CMS, e-commerce, API connections, automations and integrations between applications, databases and external services.
๐Ÿข Strategy and leadership โ€” 12 โ€” assist with decision-making, planning, organisational management, business strategy, initiative prioritisation and specialist coordination.
๐Ÿ“ Documentation and knowledge โ€” 4 โ€” create technical and process documentation, guides, knowledge bases, summaries and materials for knowledge transfer.
๐Ÿงฉ Other and universal agents โ€” 90 โ€” cover cross-cutting roles, orchestration, critique, verification, context management and tasks that do not fall within a single domain.
A4

๐Ÿงฉ Permanent domain swarms

๐Ÿ—‚๏ธ๐Ÿค 01-czechai โ€” 24 agents โ€” an in-house team for AML, banks, insurance companies, real estate, Czech law, OSINT and the day-to-day operation of proprietary services.
๐Ÿ“ฃ๐Ÿค 02-marketing โ€” 21 agents โ€” connects positioning, copywriting, campaigns, creative production, performance measurement, automation and the entire customer lifecycle.
๐Ÿ”ฌ๐Ÿค 03-vyzkum โ€” 18 agents โ€” delivers multi-source research, fact-checking, citations, competitive analysis and trend monitoring.
๐Ÿ’ป๐Ÿค 04-vyvoj โ€” 25 agents โ€” covers architecture, backend, frontend, programming languages, debugging, testing and application refactoring.
๐Ÿงช๐Ÿค 05-kvalita โ€” 17 agents โ€” performs code reviews, testing, security checks, independent peer review and final output verification.
๐ŸŽจ๐Ÿค 06-design โ€” 10 agents โ€” handles UI, UX, user research, accessibility, 3D design and visual storytelling.
๐Ÿ–ฅ๏ธ๐Ÿค 07-infra โ€” 20 agents โ€” manages servers, containers, networks, monitoring, incidents, reliability and operational security.
โš–๏ธ๐Ÿค 08-pravo โ€” 6 agents โ€” analyses contracts, compliance, GDPR, regulatory obligations and legal risk management.
๐Ÿค–๐Ÿค 09-ai โ€” 9 agents โ€” builds agent assemblies, prompts, MCP tools, orchestrations and repeatable AI workflows.
A5

โšก Task swarms

๐Ÿ•ต๏ธ๐Ÿค OSINT โ€” comprehensive background check โ€” lead osint-spy coordinates the background check of a person or company; specialists handle AML, law, search and fact-checking, and the reviewer Evidence Collector verifies evidence; the assembly utilises 21 tool agents and the services OSINT Swarm, OSINT API, OSINT Provฤ›rka, Crucix and SpiderFoot.
๐Ÿ ๐Ÿค Search swarm โ€” real estate โ€” led by offmarket-realitni-hunter; agents are engaged in parallel off-market-undervalued-hunter, realestate-agent, Investment Researcher, url-context-validator a service-auditor, while Evidence Collector verifies the accuracy and traceability of the result.
๐Ÿก๐Ÿค Reality โ€” property verification โ€” combines a real-estate agent, lawyer, AML, OSINT and investment research for due diligence on a property, its owner, rights, valuation and associated risks.
๐Ÿ•ต๏ธ๐Ÿค AML and sanctions โ€” screens a person or company against sanctions, PEP, insolvency and other risk signals, then has the results reviewed by an independent challenger.
โš–๏ธ๐Ÿค Law and contracts โ€” analyses contracts, legal risks, GDPR and compliance with Czech legislation using legal, document and privacy specialists.
๐Ÿฆ๐Ÿค Banks and mortgages โ€” brings together mortgage calculations, creditworthiness, comparison of banking products, insurance and an investment perspective into a single financial assessment.
๐Ÿ’ป๐Ÿค Web and application development โ€” guides a product from architecture through frontend, backend, TypeScript and UI all the way to an independent review of the finished code.
๐Ÿž๐Ÿค Debugging and bug finding โ€” proceeds from symptom to root cause, connecting debugging, tracing, build resolution and testing, and concludes with verification of the fix.
๐Ÿ”’๐Ÿค Security audit โ€” checks configuration, dependencies, secrets in code, architecture and potential attack vectors using auditors, penetration testers and SecOps specialists.
๐Ÿ–ฅ๏ธ๐Ÿค Infrastructure and incidents โ€” resolves downed services, outages, full disks and other operational issues from diagnosis through incident coordination to service restoration and verification.
๐Ÿ”ฌ๐Ÿค Research and desk research โ€” delivers multi-source research with search, claim verification, citations, knowledge synthesis and a final peer review.
๐Ÿ“ฃ๐Ÿค Marketing campaign โ€” guides a brief from positioning through copy, creative and paid advertising to measurement and a mandatory critical review of the campaign.
๐Ÿ“Š๐Ÿค Data and analytics โ€” connects data science, analytics, research, SQL and database optimisation for output that is verified both statistically and technically.
๐ŸŽจ๐Ÿค Design and accessibility โ€” combines UI design, UX research, visual storytelling and an accessibility audit, with the result reviewed by an accessibility testing specialist.
๐Ÿค–๐Ÿค AI, agents and MCP โ€” designs agent architecture, prompts, MCP tools, workflows and task allocation, then stress-tests them with a red-team specialist.
๐Ÿ“ฅ๐Ÿค Data collection and scraping โ€” collects data from the web, checks URL liveness, normalises results, adds context and has the evidential output independently verified.
๐Ÿ“š๐Ÿค Documentation and knowledge โ€” produces documentation, guides, cited knowledge bases and knowledge-transfer materials, then subjects them to a final critical review.
A6

๐Ÿ”Œ MCP servers โ€” connecting agents to the real world

๐Ÿ“ฆ 15 MCPs via `npx` โ€” tools are downloaded on demand from the npm registry and give agents access to specific external capabilities without a permanent local installation.
๐Ÿ–ฅ๏ธ 12 MCPs via SSH to server โ€” the agent uses a secure tunnel to services running on a remote server and works with them as available tools.
๐ŸŒ 11 MCPs via HTTP/SSE โ€” remote services communicate via a web interface and streaming events, enabling APIs, registries and custom production systems to be connected.
๐ŸŸข 7 local MCP processes โ€” tools run directly on the workstation and serve local data, files or applications.
๐Ÿ 1 local Python MCP โ€” a standalone Python process provides a specialised capability through a standard agent interface.
A7

๐Ÿ”Œ Plugins, skills, commands and automations

๐Ÿงฉ 26 plugins โ€” the installed layer includes Notion, Claude HUD, management CLAUDE.md, code review, Codex, Compound Engineering, Context7, customer support, two datasets, design, engineering, enterprise search, feature development, finance, frontend design, GitHub, legal, marketing, operations, product management, productivity, Pyright LSP, sales, Stripe and Vercel.
๐Ÿ› ๏ธ 1 009 skills in the full catalogue โ€” alphabetical groups contain: 0โ€“9 3, A 80, B 35, C 89, D 66, E 24, F 33, G 64, H 17, I 24, J 9, K 6, L 31, M 50, N 35, O 18, P 103, Q 14, R 61, S 143, T 39, U 13, V 15, W 24, X 4, Y 1 a Z 8.
โŒจ๏ธ 77 slash commands โ€” ready-made commands such as /analyze and others trigger pre-built workflows for analysis, development, testing, documentation, agent orchestration and operational work.
๐Ÿช `czechai-guardrail.js` โ€” a security hook that automatically monitors changes and enforces control guardrails when working with code and sensitive parts of the system.
๐Ÿ•ธ๏ธ `gitnexus` โ€” a hook that connects changes to repository context and helps agents understand relationships, impacts and dependencies in the code.
๐Ÿง  `herdr-agent-state.ps1` โ€” a hook that maintains and passes state across a multi-agent process so that individual roles work with a shared context and do not lose track of the task.
๐Ÿ“ `aurakit-security.md` โ€” a global security rule loaded in all sessions, protecting work independently of the specific project.
A8

๐Ÿ†• Newly added agents

๐Ÿ†• `czechai-change-reviewer` โ€” independently reviews changes after code is modified or written, evaluating correctness, security, conventions and compliance with AuraKit rules.
๐Ÿ†• `czechai-team-lead` โ€” coordinates multiple autonomous agent collaborators using a shared task list and direct messaging.
๐Ÿ†• `mrkt-automation-architect` โ€” designs marketing automation in Mautic, n8n, forms, landing pages, campaigns and dynamic content.
๐Ÿ†• `mrkt-budget-optimizer` โ€” allocates budget across channels based on expected ROI, saturation and the carry-over effect of campaigns.
๐Ÿ†• `mrkt-clv-analyst` โ€” calculates customer lifetime value and predicts future purchasing behaviour using BG/NBD and Gamma-Gamma models.
๐Ÿ†• `mrkt-competitive-intel` โ€” builds competitor profiles, identifies market gaps, tracks share of voice and prepares materials for brand differentiation.
๐Ÿ†• `mrkt-copy-chief` โ€” writes and rewrites conversion copy for websites, landing pages, ads and email campaigns.
๐Ÿ†• `mrkt-council-critic` โ€” conducts adversarial critique of marketing output before a campaign launch or copy publication.
๐Ÿ†• `mrkt-creative-studio` โ€” designs advertising creatives, marketing visuals, visual concepts and prompts for AI generation.
๐Ÿ†• `mrkt-cro-optimizer` โ€” optimises conversions on websites, forms, sign-ups, paywalls and other critical points in the customer journey.
๐Ÿ†• `mrkt-email-copywriter` โ€” prepares email sequences, newsletters, B2B cold outreach and message subject lines.
๐Ÿ†• `mrkt-growth-hacker` โ€” designs growth loops, referral programmes, free tools, directory distribution and viral mechanisms.
๐Ÿ†• `mrkt-influencer-manager` โ€” manages influencer campaigns from creator sourcing through contracts and execution to ROI evaluation.
๐Ÿ†• `mrkt-lifecycle-manager` โ€” builds email and SMS lifecycle flows, segmentation, lead scoring, retention programmes, anti-churn processes and failed payment recovery.
๐Ÿ†• `mrkt-market-researcher` โ€” conducts Voice of Customer research, reviews and interviews, and translates real customer language into marketing materials.
๐Ÿ†• `mrkt-master` โ€” acts as the primary orchestrator of the marketing fleet for a launch, a complex campaign or an entire marketing strategy.
๐Ÿ†• `mrkt-measurement-lab` โ€” designs tracking in GA4 and GTM, A/B tests, attribution, experiments and a measurement plan including hard-to-measure dark social.
๐Ÿ†• `mrkt-mmm-analyst` โ€” uses a Bayesian Media Mix Model to calculate per-channel ROAS, adstock, saturation and uncertainty intervals.
๐Ÿ†• `mrkt-paid-ads-commander` โ€” manages Google, Meta and LinkedIn campaigns including account structure, targeting, bidding, pacing and measurement.
๐Ÿ†• `mrkt-positioning-strateg` โ€” defines positioning, ideal customer profile, product category and brand narrative before content creation begins.
๐Ÿ†• `mrkt-revops-conductor` โ€” connects marketing and sales through lead lifecycle, scoring, routing, SLA and GDPR consent governance.
๐Ÿ†• `mrkt-seo-geo` โ€” increases organic visibility in traditional search engines as well as content citability in AI responses and LLM systems.
๐Ÿ†• `mrkt-social-video` โ€” creates social calendars and ready-to-use scripts for Reels, TikTok, Shorts and other short-form videos.
A9

๐Ÿ–ฅ๏ธ Workstations and server โ€” what runs where

๐Ÿ’ป Workstation โ€” when the computer is on, provides 749 agents, 1 009 skills, 77 commands, 46 MCP servers and 26 installed plugins.
๐ŸŒ Server 179 โ€” operates independently even without the workstation and, according to measurements from 3. 8. 2026, contains 731 agents, 979 skills, 77 commands and 34 MCP servers.
๐Ÿงฉ Plugins on workstation only โ€” the plugin layer is not installed on Server 179; remote operation uses server-side agents, skills, commands and MCP connections.
๐Ÿ–ฅ๏ธ Claude Code 2.1.220 on Server 179 โ€” the stated version forms the runtime environment of the remote Arsenal instance.
A10

๐Ÿงฑ Internal organisation of the Arsenal

๐Ÿชช Each agent has its own system ID โ€” the record contains a professional category, a Czech name, a brief description of the capability and the system identifier used during actual execution.
๐Ÿ—‚๏ธ Categories serve navigation, IDs serve execution โ€” the professional classification facilitates selection, while the orchestrator and swarms work with verified identifiers of specific agents rather than generic role names alone.
๐Ÿค A swarm has a clearly defined responsibility โ€” the configuration distinguishes a lead, parallel specialists and a challenger; for selected teams it also records connected services, tool agents and operational endpoints.
๐Ÿ› ๏ธ A skill is loaded only when used โ€” the library can contain more than a thousand skills, because individual instructions do not occupy the working context until a specific agent actually needs them.
๐Ÿ”Œ MCP is registered according to its runtime method โ€” the catalogue distinguishes tools launched via npx, SSH tunnel, HTTP/SSE, local process, and local Python service.
๐Ÿ–ฅ๏ธ The workstation and server have separate inventories โ€” The Arsenal precisely distinguishes which agents, skills, commands, plugins and MCPs the workstation has and which are independently operated by server 179.
A11

๐Ÿšจ Iron operational rules

๐Ÿ‘๏ธ Visual verification before marking as "done" โ€” an HTTP 200 server response alone is not sufficient; the result must actually be displayed and verified, for example by a screenshot.
๐ŸŒ Scraping and OSINT from servers only 134/179 โ€” public data collection and OSINT operations are not performed directly from the workstation.
๐Ÿ”ด AML core and production website only upon confirmation โ€” the sensitive core and live website are not touched without explicit consent.
๐Ÿง  Memory reflex โ€” at the start of work, relevant memory is loaded, and at the end, new findings and decisions are saved.
A12

๐Ÿ“ Note on inventory accuracy

๐Ÿ“… Two revision dates โ€” the summary inventory of unique items is marked on the website with the date 7. 8. 2026, while the capabilities overview, operational status and some lower blocks use the date 3. 8. 2026.
โš ๏ธ Differing totals on a single page โ€” the upper overview lists 23 categories, 26 swarms and 44 MCPs, the detailed sections list 26 swarms and 46 MCPs, and the older lower blocks still display 22 categories and 41 swarms; this document therefore uses the detailed breakdown of 23 categories, 26 swarms and 46 MCPs and explicitly preserves the discrepancy.
๐Ÿงฎ Sum of professional cards โ€” the visible counts across 23 professional categories add up to 708 items, while the header records 749 entries and 687 unique agents; these figures describe different layers of the inventory and should not be confused with one another.
Source: arzenal.czechai.io ยท live catalogue verified 23 Aug 2026. This section is maintained in the master as a standalone, non-duplicate technical layer separate from customer-facing services.
05

๐Ÿ“Š IV. VERIFIED METRICS, DATA AND INFRASTRUCTURE

ฤŒ.MetricValueMeasurement source
M1โฑ๏ธ Continuous development and operations8 years (2018โ€“2026)brand history
M2๐Ÿงฉ Proprietary applications, modules and outputs1 550+conservative count; the public directory alone contains 6,521 unique URLs across 110 folders
M3๐ŸŒ Live subdomains on czechai.io130 (HTTP 2xx)asset scan 2 Aug 2026
M4๐ŸŒ Live hostnames across domains153asset scan 2 Aug 2026
M5โš™๏ธ Domains and subdomains in nginx configuration138production server configuration
M6๐Ÿ”— Root domains held16export from registrars
M7๐Ÿ–ฅ๏ธ Running PM2 production services1 500 across six serverspm2 jlist per server
M8๐Ÿณ Docker containers273 runningdocker ps -q per server
M9๐Ÿค– AI agents and specialised teams4 020proprietary registry of agents, swarms and skills
M10๐Ÿ•ธ๏ธ Of which in the production registry73 in 40 swarms/v1/agents/registry/stats
M11๐Ÿ“š Of which in the local definition library749 agents + 1 475 skillsworkstation inventory
M12๐Ÿ› ๏ธ Shared tools for agents15agent registry
M13โš–๏ธ Court decisions for Legal AI578 032justice_decisions.db, table decisions
M14๐Ÿ“– Legal regulations and sections17 679 regulations / 30 712 sectionsczech_laws.db
M15๐Ÿง  Vector knowledge base โ€” case law617 362 vectorsQdrant judikatura_384
M16๐Ÿ“š Vector knowledge base โ€” e-Sbรญrka277 644 vectorsQdrant esbirka_384; 23 collections in total
M17๐Ÿ›ก๏ธ Sanctions records in the AML layer3.6 mil.AML platform; live API currently holds 1 689 782 in memory
M18๐Ÿข Enriched ARES company records3 255 092ares_enriched.db, table companies
M19๐Ÿฆ AML layer modules18AML platform
M20๐Ÿ“„ HTML pages in public production4 586/var/www/router-static, 25 GB
M21๐Ÿ–ฅ๏ธ Proprietary servers6 + GPU server RTX PRO 6000, 96 GB VRAMinfrastructure
M22๐Ÿ“ˆ Apartment valuation accuracy โ€” documented backtest V24Rยฒ 0,948proprietary backtest V24; current product line is V26
Note: inventories were created across different layers and at different points in time. This is why, for example, the aggregate 4 020 agents, teams and skills, local data 749 agents + 1 475 skills and Arsenal detail 749 records, 687 unique agents and 1 009 skills do not automatically exclude each other, but must not be presented as identical metrics without explanation. Before any public or investment use, it is advisable to re-confirm each value against the stated source.
06

๐Ÿ”— V. MAIN PUBLIC PLATFORMS AND OUTPUTS

๐ŸŒ CzechAI โ€” https://czechai.io/
๐Ÿ“˜ Complete whitepaper โ€” https://whitepaper.czechai.io/
๐Ÿญ CzechAI Strojovna โ€” https://strojovna.czechai.io/
๐ŸŽ“ CzechAI Academy โ€” https://academy.czechai.io/
๐ŸŽฌ Demos and case studies โ€” https://show3.czechai.io/
๐Ÿ  EstateX โ€” https://estatex.club/
๐Ÿค– EstateX Copilot โ€” https://copilot.czechai.io/
๐Ÿ“ˆ Price Map V26 โ€” https://shadow.czechai.io/
๐Ÿฆ HypoStars โ€” https://hypostars.cz/
๐Ÿ’ฌ CzechAI Chat โ€” https://chat.czechai.io/
โš–๏ธ Prรกvnรญk AI โ€” https://law.czechai.io/
๐Ÿงญ CzechAI Portal โ€” https://portal.czechai.io/
๐Ÿ” AI Rentgen โ€” https://rentgen.czechai.io/
๐Ÿ–จ๏ธ Digital Print Shop โ€” https://tiskarna.digital/
๐Ÿ›’ Digital Marketplace โ€” https://trziste.digital/
๐Ÿ“ฌ Rozesรญlka.eu โ€” https://rozesilka.eu/
๐ŸŒ Worldmonitor โ€” https://monitor2.czechai.io/
๐Ÿงฌ Arzenรกl โ€” https://arzenal.czechai.io/
07

๐Ÿ“ VI. RULES OF THIS MASTER

One piece of information, one primary location ๐Ÿ“Œ โ€” services are in Part I, cross-cutting technologies and development in Part II, the internal agent inventory in Part III, metrics in Part IV, and public links in Part V.
No loss of content โœ… โ€” shorter and longer variants were merged into the most complete formulation; only genuine or semantically identical duplicates were removed.
Preserved names and parameters ๐Ÿท๏ธ โ€” product names, domains, technologies, counts, revision dates, differing inventory totals, and notes on their methodology are all retained.
Separation of offering from internal tooling ๐Ÿงฌ โ€” Claude Arzenรกl is not presented as a standalone customer-facing service; it is the operational and tooling layer through which services are delivered.
Verification before publication ๐Ÿ” โ€” dynamic counts, operational status, regulatory claims, and designations such as โ€œlargestโ€ must be confirmed by current measurement or a verifiable source before any public publication.
08

โœ… VII. COMPLETENESS CHECK

โœ… Services and products โ€” 19 numbered categories and 120 individual customer items.
๐Ÿš€ Technologies and development โ€” 9 separate technology categories and 161 detailed records without merging into brief summaries.
๐Ÿงฌ Claude Arsenal โ€” preserved Claude Stack introduction and complete inventory of professions, permanent swarms, task swarms, MCP servers, plugins, skills, commands, hooks, new agents, operational deployment and iron rules.
๐Ÿ“Š Metrics โ€” 22 numerical indicators including source or measurement method and explanation of differing inventory totals.
๐Ÿ”— Public Outputs โ€” 18 specific links to live products and platforms; technical router crossroads and currently non-functional targets are not listed.
๐Ÿ˜€ Emoji Check โ€” every category heading and every registered record, service, agent item, metric, rule and public link contains its own emoji.
๐Ÿ”Ž Name Check โ€” automatic comparison against the master service catalogue and the standalone Arsenal catalogue found no missing source names.
๐Ÿ“Œ Canonical file designation: the single master catalogue of the CzechAI ecosystem. Replaces the need to use the compact service catalogue, the standalone Arsenal catalogue and unstructured lists of technology or development areas in parallel.
GrafitPapรญrKonzole