Software Engineer · Backend & AI-Assisted Systems
I build backend and data systems in which authorization, correctness, and predictable failure handling matter as much as feature delivery.
At Varahe Analytics, I work across Python services, API integrations, data platforms, and evidence-aware AI workflows. I completed my B.Tech in Computer Science and Engineering from IIIT Nagpur with a 9.97 CGPA.
- Backend systems: Python, FastAPI, Django, Node.js, REST APIs, OAuth2/JWT, and third-party integrations.
- Data platforms: PostgreSQL, Redis, ETL, synchronization, identity resolution, audit trails, and governed access.
- Reliable AI workflows: retrieval, embeddings, structured outputs, agent orchestration, schema validation, evidence provenance, retries, and fallbacks.
- System design: explicit sources of truth, temporal state modeling, least-privilege authorization, observability, and recovery paths.
Most of my production work is private; these descriptions intentionally omit proprietary implementation details.
- Secure data-access gateway — Architected a FastAPI/PostgreSQL service that exposes policy-scoped HR data to approved clients through OAuth2 client credentials, revocable JWTs, field-level authorization, selective AES-256-GCM decryption, rate limiting, and auditable access.
- Rule-driven operations platform — Owned backend rules, REST APIs, and Slack integrations for a platform used by approximately 1,000 consultants; modeled historical task state through sparse temporal rules rather than dense daily snapshots.
- Governed data and AI workflows — Built normalized data pipelines and evidence-aware reporting systems that combine deterministic controls with bounded model decisions, schema validation, retries, fallbacks, and deterministic publishing.
- ApniDukaan — Built an offline-first Flutter/Dart ledger for customer and supplier credit management, with atomic balances, settlements, reports, OTP/Google authentication, and collaborator access. It is used in two real-world contexts, including the daily operations of a medical retailer.
- TalkNest — Built and deployed a browser meeting prototype using Socket.IO for room coordination and WebRTC for peer-to-peer media, with screen sharing, chat, file exchange, and participant controls. Live application.
| Area | Technologies |
|---|---|
| Languages | Python, SQL, JavaScript, C++, Java |
| Backend | FastAPI, Django, SQLAlchemy, Pydantic, Node.js, Express.js |
| Data | PostgreSQL, MySQL, Redis, BigQuery, MongoDB, Qdrant, Firestore |
| AI systems | LLM APIs, embeddings, structured outputs, agent orchestration, validation and guardrails |
| Platform | GCP, Cloud Run, Cloud SQL, GCS, Docker, Git, Linux, CI/CD |
- B.Tech, Computer Science and Engineering — IIIT Nagpur, 9.97 CGPA
- LeetCode Knight · CodeChef 4-star, maximum rating 1773 and Global Rank 205 in Starters 104
- Winner — Hack-o-Mania, IIT Ropar · Winner — Hackswift International Hackathon
