name Abdulaziz Abdujalilov, aka Balu
role AI Specialist + Full-Stack Developer + LLMOps + Digital Media Strategist
style useful AI systems, clean interfaces, evals, observability, reliable releases
mission build tools that save time, explain ideas clearly, and survive production
location Tashkent, UzbekistanI like projects where code is not just code. The best work has three parts: a useful system, a clean product experience, and a story people can understand fast. The DevOps layer matters because AI products need reliability, observability, cost control, evals, and rollback plans.
|
Desktop product. Embedded Gemma 4 via Ollama, glassmorphism UI in Russian and Uzbek. Nine sections: chat with conversation history, email drafts with tone control, long-text summarization, RU/UZ/EN translation, grammar and rewrite, unit/currency/file-size converters, RU and Uzbek transliteration. First-run wizard auto-detects the engine, streams model download with live MB/s and ETA, and supports skip-and-install-later for offline users. |
|
|
Mini-SaaS campaign system with provider mode, local campaign library, prompt packs, roadmap, review gates, and exports. |
Prompt regression tests with datasets, deterministic mock model, quality scoring, and CI gates. |
|
Knowledge base agent with chunking, mock vector search, cited answers, Docker-ready structure, and tests. |
Agent workflow engine with planner, tools, memory, human approval gate, retries, and execution logs. |
|
CI-ready release guard for AI services: health, budget, rollback, observability, and model policy checks. |
LLMOps metrics for latency, token usage, model cost, success rate, and error budget burn. |
Production-style RAG deployment blueprint with Docker, health checks, config validation, and runbooks. |
+ Building AI-powered tools for creators and small businesses
+ Turning content strategy into repeatable systems
+ Shipping full-stack demos with clean logic and live previews
+ Adding AI evals, RAG, agents, LLMOps, observability, Docker, and release safety| AI Prompt Engineering Playbook Prompt patterns, schemas, critic loops, and product-ready AI workflow structure. |
LLM Evaluation Checklist A practical quality checklist for testing AI features before shipping. |
| RAG System Blueprint Architecture notes for retrieval, chunking, metadata, citations, and monitoring. |
AI DevOps CI/CD Template CI/CD structure for AI apps, prompt tests, eval smoke checks, and release safety. |
| Digital Media AI Growth System How AI, content strategy, analytics, and full-stack tooling can work as one growth system. |
|
