I'm Piku Maity — an AI Engineer based in Bangalore, India.
I build production-grade AI systems that solve real business problems — not demos, not notebooks, not toy projects. Over the past 3+ years, I've shipped AI solutions at CSG Karnataka and Philips across Healthcare, Transport, Enterprise, and Public Sector domains.
My core expertise sits at the intersection of GenAI, RAG pipelines, AI Agents, LLMs, and MCP — the full stack of what it takes to move an AI idea from whiteboard to production.
Outside of engineering, I run an AI community of 30K+ builders and engineers across LinkedIn, YouTube, and Twitter — teaching production AI for those who want to actually ship, not just follow trends.
🎓 RV College of Engineering, Bangalore | 💼 AI Engineer @ CSG Karnataka
"The difference between an AI experiment and an AI system is whether it survives contact with real data, real users, and production constraints. I build for the latter."
- 🏗️ Building → Web Audit Agent v2 — AI-native DAST Platform (open-source Burp Suite / ZAP alternative)
- 🌱 Exploring → MCP (Model Context Protocol) | Advanced RAG Pipelines | LLM Fine-tuning
- 🚀 Deploying → Open-weight LLMs on NVIDIA H100/H200 | VRAM optimization | Quantization
- 📝 Creating → Production AI breakdowns on GenAI, AI Agents, RAG, and LLM deployment — for engineers who ship
- 🧩 I solve → RAG that doesn't hallucinate · Agents that don't loop · LLMs that run on your hardware
- ⚡ Fun fact → My AI models handle edge cases better than my jokes handle silence!
| 🤖 Agentic AI & LLMs | 🔍 RAG & Search | 🏗️ Production Systems | 🔧 AI Infra & Inference |
|---|---|---|---|
| Multi-Agent Orchestration | Semantic Search | FastAPI + Docker | vLLM · SGLang · TGI |
| LangChain & LangGraph | Vector DB Design | GPU Inference (H100/H200) | Ollama · Open WebUI |
| MCP Integration | Hybrid RAG Pipelines | LLM Serving & Deployment | NVIDIA H100 / H200 |
| Tool Calling & Memory | Embedding Models | PostgreSQL + DuckDB | VRAM Optimization · Quantization |
| Structured Output & Parsing | ChromaDB / Qdrant | CI-ready AI Pipelines | GGUF · AWQ · GPTQ |
| Prompt Engineering | Context-Aware Retrieval | Python · REST APIs | Docker · Linux · Git |
Production-grade AI systems shipped across Healthcare, Transport, Enterprise & Public Sector.
| 🔧 System | 🏭 Domain | 🛠️ Stack | 📈 Impact |
|---|---|---|---|
| 🔐 Web Audit Agent — AI-Native DAST Platform | Security | Python, Multi-Agent, Playwright, 5 LLM Providers | Open-source Burp Suite/ZAP alternative — 16 scanners, 8 AI personas, 47 core modules |
| 📄 DPR CoPilot — Document Intelligence | Enterprise | RAG, LLMs, Vector DB, Document Parsing | Reduced document lookup time significantly — instant Q&A over large enterprise doc repositories |
| 🚌 KSRTC AI Systems — Operational Analytics | Public Sector | ML, NLP, PySpark, BI Dashboards | Automated reporting and analytics for Karnataka's state transport ops — replaced manual processes |
| 🏥 Health Insurance ML App | Healthcare | KNN, Random Forest, Logistic Regression | 60%+ improvement in policy renewal prediction accuracy |
| 🖥️ Enterprise Desktop Onboarding App | Enterprise | Python, PyQT5, PySide2, MySQL | 40% productivity boost for Data Migration teams across enterprise onboarding workflows |
| 💼 Service | 📋 Description |
|---|---|
| 🔬 AI Consulting | Architecture design for GenAI, RAG & Agentic AI systems |
| 🛠️ Freelance AI Projects | Production AI system development — agents, RAG pipelines, LLM deployment |
| 🎤 Speaking & Advisory | AI/ML talks, webinars, workshops, panel discussions |
| 🤝 Brand Collaborations | Sponsored content, product reviews, co-creation (30K+ AI community) |
| 🏢 Full-time Opportunities | Senior AI Engineer / AI Architect roles focused on GenAI & Agentic AI |
Found something useful? Drop a ⭐ — it keeps me shipping. Always open to collaborate.