I build AI systems that ship to production. 12+ years of engineering experience spanning embedded systems, cloud infrastructure, and AI — currently part of the Data and AI team at Red Hat, focused on agentic AI, RAG pipelines, and LLM deployment.
- 🔬 AI Researcher — Two papers submitted to NeurIPS 2026 (AmbiguityBench + cross-provider LLM behavioral audit)
- 💡 Patent Inventor — 9 patents in generative AI, intelligent agents, and data security
- 🎤 Conference Speaker — Delivered talk at DevConf.IN 2026, Pune on production RAG systems; talk selected at DevConf.CZ 2026
- 📝 Technical Writer — 30+ articles on Medium · 26K+ reads · Topics: RAG, agentic AI, LLM observability, transformers
| Project | Description | Link |
|---|---|---|
| AntarDarshan | Production RAG system over 54 classical Indian philosophy texts — hybrid retrieval (BGE-M3 dense + sparse), cross-encoder reranker, Qdrant, FastAPI, Next.js | antardarshan.org |
| Library | Description | Install |
|---|---|---|
| chunking-strategy | Production-grade semantic text chunking — thread-safe, streaming, adaptive retrieval feedback | pip install chunking-strategy |
| llm-smartmem | Smart memory management for LLM conversations — topic-aware compression for agentic systems | pip install llm-smartmem |
Agentic AI & LLMs
ML / DL
Infrastructure
- Agent Skills: The Quiet Standard That's Changing How We Build AI Agents
- The Complete Guide to Embeddings and RAG: From Theory to Production — 7.2K reads
- Embeddings: A Deep Dive from Basics to Advanced Concepts — 6K reads
- The Rise of MCP: How a "USB-C for AI" Is Reshaping Intelligent Systems
- Understanding Mixture of Experts (MoE)
-
🎤 DevConf.IN 2026, Pune — "Why Your RAG System Hallucinates: Fixing the Content Segmentation Problem" · Watch on YouTube · Talk Details · Demonstrated 40–60% retrieval accuracy improvement live
-
🎤 DevConf.CZ 2026, Czech Republic — "Beyond Token Limits: Building Memory That Actually Works for LLM Agents" · Related library: llm-smartmem
Want to connect or book a mentoring session?
