Architect & Delivery Lead with 13+ years in software engineering, now building Agentic AI and LLM systems that hold up in production.
By day I lead architecture and delivery for large-scale microservices and micro-frontend platforms (.NET, Python, Angular, Azure). By night I build AI agents, RAG pipelines and learning tools, and explain them simply on YouTube.
| Project | What it is | Stack |
|---|---|---|
| Locus | Local-first RAG research workspace: upload files, ask questions, inspect the retrieval pipeline live | React 19 · FastAPI · pgvector · Ollama / Groq / OpenAI / Gemini |
| AgentArmy | Multi-agent task pipelines: a Commander LLM delegates to worker agents, each backed by its own MCP server | Python · TypeScript web UI · MCP · YAML pipelines · multi-provider LLMs |
| LKG | .NET Architect → GenAI Architect in 80 reel-style lessons with voiceover, captions and quizzes | PWA · vanilla JS · TTS |
| Myna | Hyperlocal "what's near me" finder for street vendors and local shops no app lists | Python |
Python .NET Core Azure Azure DevOps Microservices Micro-frontends LLMs Agentic AI MCP RAG Claude / Anthropic SDK
- YouTube: Kwikwise — AI agents & system design, explained in minutes
- LinkedIn: linkedin.com/in/realsharad

