I am a software engineer focused on building low-latency backend systems and production-ready Artificial Intelligence pipelines. I specialize in Machine Learning, robust LLM integrations (RAG), and writing extensively tested, scalable code. I actively debug and contribute to enterprise-grade open-source repositories to improve system reliability.

• Languages: Python, Java, SQL
• Machine Learning & AI: Machine Learning, Deep Learning, PyTorch, LLMs, RAG, LangChain, LlamaIndex
• Backend Infrastructure: FastAPI, Flask, REST API Design, Vector Databases (Pinecone/Chroma), PostgreSQL
• Cloud & DevOps: AWS, Docker, Git, GitHub Actions, CI/CD, Unit Testing (Pytest, Unittest)

Enterprise RAG Document Search System

• Domain: Applied AI & Backend
• Impact: Engineered an AI-powered Retrieval-Augmented Generation (RAG) system capable of querying and extracting precise context from 10,000+ dense documents with sub-100ms latency.
• Tech: Python, LangChain, Vector Databases, Machine Learning.

Guardian Gate System Architecture

• Domain: Backend & System Design
• Impact: Designed a highly scalable backend architecture focusing on secure request routing, API gateways, and efficient load management to handle high-throughput network traffic.
• Tech: Backend Architecture, API Security, System Design.

LangChain (langchain-ai)

• Engineering: Resolved a critical crash in ChatOllama caused by malformed JSON during tool calling streams. Wrapped the async generator to catch ResponseError, allowing the agent loop to gracefully recover and process raw text instead of hard-failing.
• Testing: Validated recovery logic with comprehensive pytest suites.
• Status: Active PR (#35201)

AWS CLI (aws)

• Engineering: Prevented configuration corruption by modifying the config writing logic to skip empty or whitespace-only inputs (e.g., region =). Preserved expected behavior for valid inputs across the CLI ecosystem.
• Testing: Authored rigorous unit tests adhering to strict AWS unittest standards.
• Status: Active PR (#10003)

LlamaIndex

• Engineering: Investigated Mem0 memory integration behavior with multiple context identifiers. Proposed architectural adjustments during technical reviews with core maintainers, driving an upstream system refactor.

TheAlgorithms/Java (~65k Stars) & keon/algorithms (~25k Stars)

• Engineering: Implemented foundational algorithms (Euclidean Distance, Torus Surface Area) in Java and Python. Ensured strict adherence to CI/CD pipelines, clang-format standards, and mathematical correctness.
• Status: 3 PRs Merged (#7218, #2713, #2714)