Praktika app
Three original programmes, daily practices and visual progression. Programmes, a journal and a practice library.
Explore the app →Working systems · Prototypes · Physical results
AI that's
in real projects.
Six tracks where neural networks, RAG and agents have already helped build a product, automate a process or assemble a verifiable prototype.
WORKINGin active use
PROTOTYPEarchitecture and user scenario validated
PRODUCTpublic digital or physical product delivered
Multi-channel publishing
One agent prepares content and manages connectors for VK, Telegram, LinkedIn, websites, VK Video and YouTube.
Case study →WORKING · MULTI-AGENTDigi Anton
Multi-agent system with separate roles, shared memory, local models, schedules, control and redundancy.
Project architecture →PROTOTYPE · FINANCE RAGIslamic finance RAG
A trusted knowledge base and a prototype AI consultant for partner (Islamic) finance, law and banking products.
Professional context →PROTOTYPE · LEGAL AILegal assistant
Search and answer generation from legal and Sharia sources backed by a knowledge base, not model memory.
Related project →PRODUCT · BOOKBook created with AI
A practical guide on neural networks and AI agents went from research and structure to printed product and sales.
About the book and where to buy →PRODUCT · WEB + VISUALWebsites, video and images
AI is used in research, structure, copy, visual materials and the production workflow of digital products.
Digital production →Core principle
Source, tool, control
Source — the system works with documents, data, media and accepted project memory.
Tool — the agent not only writes a response but calls a secure connector, search, generator or production scenario.
Control — status, log, result verification and a clear boundary between automation and human decision.
Laboratory · projects in development
From a prototype to an original product
A brewery management game
Production, beer styles, sales and a tasting hall in an original management game.
Explore the game →