Projects

=== Finch – News application ===

• Built an end-to-end CI/CD pipeline from source code to production deployment, implementing Docker multi-stage builds, image registry workflows, unit testing, Kubernetes deployments, and automatic scaling using HPA.
• Established a secure RDS connection using Cloud SQL Auth Proxy, enabling encrypted communication between GKE workloads and the database.
• Implemented observability using Prometheus and Grafana with custom alerts to track cluster health, performance, and application behavior.
• Successfully predicted potential production failures and proactively resolved issues using monitoring insights, improving system stability and reliability.

=== DocuMind – AI-Powered Document Q&A System ===

• Developed a FastAPI-based backend that enables users to upload PDFs or text files, extract content, and perform intelligent question-answering using OpenAI GPT-4 and LangChain workflows.
• Integrated Pinecone for embedding storage and Supabase for metadata, automating document ingestion, chunking, and retrieval pipelines, resulting in a 60% improvement in contextual response accuracy.
• Containerized the backend using Docker and deployed it on Render with GitHub Actions–based CI/CD for seamless version control and scalability

=== VoiceQuery – Voice-Driven Knowledge Assistant ===

• Built an interactive AI assistant using FastAPI that combines OpenAI Whisper for speech-to-text, GPT-4 for reasoning, and ElevenLabs TTS for generating natural spoken responses.
• Integrated LangChain for contextual prompt flow management and Chroma DB for semantic vector storage, enabling low-latency conversational retrieval across voice queries.
• Deployed the complete pipeline on Render using Docker and implemented CI workflows for continuous delivery and version tracking.

=== QA Rag Pipeline ===

A Simple Rag Pipeline Using Langchain and chromavectordb , Ollama ,HuggingFace Embedding Models github

Architecture

image