I'm an AI Full Stack Developer and a 2026 Computer Engineering graduate, currently working at NeuralArc.ai in Pune. I build backend services and AI-powered features, mostly with Python, FastAPI, and React, and I enjoy the part of the job where LLMs, APIs, and databases have to work together reliably in production.
- 💼 Working on production AI systems at NeuralArc.ai, including ModelBeat.ai (multi-model inference platform)
- 🔍 Interested in RAG pipelines, model orchestration, and real-time voice/audio
- ⚙️ Care about clean APIs, query performance, and debuggable backends
- 🌱 Always learning, currently going deeper into LLM infrastructure and observability
🧠 AI Full Stack Developer · NeuralArc.ai
Feb 2026 – Present · Pune, India
- Built Python backend services with FastAPI and AsyncIO connecting to multiple LLMs via AWS Bedrock and Gemini, powering data discovery and structured data generation.
- Contributed to ModelBeat's multi-model orchestration and inference API: auto-routing, onboarding workflows, and observability for inference requests.
- Worked on parts of a RAG pipeline: document parsing (Qwen-VL, OpenDataLoader), chunking, embeddings, and vector retrieval.
- Helped integrate LiveKit and Deepgram for real-time speech-to-text and live communication features.
- Contributed to REST APIs and background workflows for multi-tenant apps with role-based access control.
- Improved backend performance with database indexing, query tuning, and Valkey (Redis-compatible) caching; helped with Stripe billing logic for subscriptions and proration.
Oct 2025 – Jan 2026 · Remote
- Worked on a RAG-based system for customer queries, feeding relevant context into AI-assisted responses.
- Contributed to backend optimization and API integrations.
| Project | What it does | Stack |
|---|---|---|
| PitchDekh | AI-powered pitch intelligence platform. Parses startup pitch decks, extracts structured data with Gemini, and supports startup-investor matching. Includes a Firecrawl pipeline for investor profile discovery. | React · Node.js · Express · Gemini API · Supabase · Firecrawl |
| RAG Chatbot | Chatbot on open-source LLMs via Groq's low-latency API, with Pinecone-based retrieval, tool-calling, and web search. | React · Node.js · Groq · Pinecone |
| Adverra | Multi-modal generative AI workflow combining LLMs, text-to-speech (including multi-speaker audio), and image generation. | React · Node.js · Gemini API · Supabase |
📌 Repo links are on my GitHub profile.
Languages
Backend & APIs
AI Engineering
Frontend
Databases
Cloud & Tools
- B.E. Computer Engineering, Dhole Patil College of Engineering, Pune (2022 – 2026) · CGPA 8.03 / 10
- MERN Development, Apna College
I'm open to conversations about AI backends, LLM tooling, and full-stack work. Feel free to reach out via email or LinkedIn.