I am a Python developer focused on Backend Engineering and AI. I build reliable APIs using FastAPI, manage data with PostgreSQL, and build tools that use AI to solve real problems
| Domain | Stack |
|---|---|
| Backend Engineering |    |
| Database & ORM |   |
| Applied AI |    |
Automated RAG-Based Screening System An intelligent document-parsing and evaluation API that deterministically scores candidates against job descriptions.
- Core Problem: Manual resume screening is slow, highly subjective, and standard keyword parsers often fail to understand true technical context.
- Solution: Architected a Retrieval-Augmented Generation (RAG) pipeline utilizing ChromaDB for semantic search, and implemented Instructor/Pydantic to force the LLM into outputting strict, hallucination-free JSON decisions (Match/Reject).
- Key Tech: Python, FastAPI, ChromaDB, Groq LLM, RAG, Streamlit.
2) BookFast API
High-Concurrency Ticketing System A robust booking API designed to handle race conditions during high-demand events.
- Core Problem: Preventing "double-booking" when thousands of users hit the endpoint simultaneously.
- Solution: Implemented Atomic Transactions and database locking strategies to guarantee inventory consistency.
- Key Tech: FastAPI, PostgreSQL, JWT Authentication, Docker Compose.
Structured Data Extraction Pipeline A vision-based tool that transforms unstructured receipt images data into strict, validated JSON.
- Core Problem: Manual data entry from receipts is time-consuming and prone to typos/human error.
- Solution: Built an intelligent extraction pipeline using Llama Vision to read messy images and Pydantic to enforce strict accuracy (ensuring data format is correct).
- Key Tech: Llama Vision (via Groq), Python, Pydantic.
I am currently transitioning from standard RAG pipelines to building autonomous, stateful AI agents.
- Active Experiment: Developing a multi-tool autonomous agent using LangGraph, Python (AsyncIO), and local SQLite persistence.
- Goal: Building an end-to-end "Coder Agent" capable of dynamic web searching, real-time web scraping, and local file system execution.
