Backend & AI/ML engineer who ships production systems — I built QuillFlow, a deployed, multi-tenant Agentic RAG platform (FastAPI · LangGraph · Qdrant) with hybrid retrieval, reranking, and citation-grounded answers.
_{📍 Open to Backend / Systems & AI / ML Engineer roles}

"Build systems that are simple to reason about and hard to break."

• 🔭 Software engineer working across backend systems and applied AI/ML.
• 🚀 Shipped QuillFlow — a production Agentic RAG platform: hybrid retrieval (dense + BM25 + cross-encoder rerank), LangGraph orchestration, multi-tenant auth, 30+ REST endpoints, and a two-tier Redis cache.
• 🤖 Hands-on with LLMs & GenAI, RAG pipelines, NLP/CV, and serving ML behind real production services.
• 💬 Ask me about Python, Java, C++, Go, designing clean scalable APIs, and putting AI into production.
• ⚡ I enjoy untangling performance bottlenecks and designing systems that are simple to reason about and hard to break.
• 📫 Open to SDE & AI/ML Engineer opportunities — reach me through the links above.

Languages

Frameworks & Backend

DevOps & Cloud

AI / ML

Applied-AI work I've shipped and the areas I'm actively pushing into:

• 🤖 LLMs & GenAI — building on OpenAI/LLM APIs, prompt engineering, RAG pipelines, and agentic workflows with LangChain.
• 📝 NLP & Computer Vision — text understanding/extraction and vision tasks with OpenCV and transformer models.
• 🚢 MLOps & Data — data wrangling with Pandas/NumPy and deploying ML behind production services.
• 📚 Core ML & Deep Learning — deepening fundamentals (PyTorch/TensorFlow) and applying them in hands-on projects.

🪶 QuillFlow — Production Agentic RAG Platform A deployed, multi-tenant RAG system that answers questions over your private knowledge base with citation-grounded responses. Built end-to-end: ingestion → retrieval → generation → serving. Hybrid retrieval — dense (Qdrant/HNSW) + sparse (BM25) + cross-encoder reranking Agentic orchestration — LangGraph DAG routes simple vs. complex queries, plans & writes multi-section answers in parallel Enterprise backend — JWT + API-key auth, RBAC, multi-tenant isolation, 30+ FastAPI endpoints Performance — two-tier Redis cache (exact + semantic), async ARQ workers, SSE streaming Python 3.12 · FastAPI · LangGraph · Qdrant · PostgreSQL · Redis · React · Docker

_{🚧 A second AI-focused project is in the works — it'll be featured here once it ships.}

"The best way to predict the future is to build it." — Alan Kay

Thanks for visiting — feel free to reach out! 🚀