Python backend developer with 7+ years of experience. Enthusiastic about building products and understanding how systems work. AI has enabled me to think beyond the code - across architecture, product, and execution - and develop a systems thinking mindset.
Work Highlights
- Built JWT-based source verification using ES256 algorithm to secure cross-app access for 40,000+ students.
- Built exam city edit module serving 37,000+ students across 3 programs, eliminating manual support team intervention.
- Added Celery-based async processing to settlement service, enabling it to handle peak loads of 8,000+ customers.
- Drove settlement success rate to 100% by reducing merchant escalations to zero and coordinating with Razorpay and other dependent internal tech teams.
Side Projects
- Prepit 2026
RAG-based knowledge base for interview prep -built end to end.
- Batch ingestion silently fell back to keyword search with no error -caught only via shell experiments.
- Deleting messages to manage context window cost meant losing training data -switched to session IDs instead.
- Hybrid search score can't measure semantic relevance -used near_vector distance separately for out-of-scope detection.
- SystemsFeed 2026
AI reading layer on top of engineering blogs -summaries, tags, prerequisites, and ELI5.
- Built tag normalization in layers: evals to measure quality, prompt refinement, semantic search for fuzzy duplicates, DB unique constraint for exact ones.
Tech Stack
- Languages
- Python, JavaScript
- Frameworks
- Flask, FastAPI
- Databases
- PostgreSQL, MySQL, Google Datastore, Redis, Weaviate
- Cloud / Deployment
- AWS EC2, Digital Ocean, Netlify, GitHub Actions
- Tools
- Git, Docker, Celery, Claude Code, Gemini CLI
- API Integrations
- OpenAI, Claude API, Slack, Razorpay, WebEngage, Gmail
- AI Engineering
- LLM Application Development, Agents, RAG Systems, LangGraph
Speaking & Content
- Conferences
- PyData Delhi, EuroPython, PyLadies meetups
- Content
- LinkedIn Learning, Twilio, egghead.io