I design and engineer AI systems that move beyond experimentation and operate reliably at production scale.
B.Tech CSE (AI/ML) at IIIT Nagpur
Focused on LLM systems, multimodal intelligence, and scalable AI infrastructure
I work across the full lifecycle:
data → modeling → training → inference → deployment
- Designing LLM and agent-based systems with structured reasoning and tool integration
- Building retrieval-augmented and graph-based pipelines
- Optimizing inference systems for latency, throughput, and cost
- Developing backend systems for AI applications with production constraints
- Running structured experimentation and evaluation for model improvement
- LLMs & Agent Systems → RAG, LangChain, LangGraph, AutoGen
- Computer Vision → Detection, tracking, multimodal systems
- Systems Engineering → Distributed systems, async pipelines
- Inference Optimization → Batching, quantization, GPU efficiency
| Achievement | Outcome |
|---|---|
| CVPR Workshop – AUTOPILOT VQA | Ranked #1 globally on multimodal evaluation benchmark (661 videos, 25 QA pipelines) |
| IEEE WCCI 2026 | Published research on multi-modal behavioral detection of malicious npm packages |
| ICMLDE 2025 | Published "QFedAstro: Quantum-Enhanced Federated Learning" in Procedia Computer Science |
| IIT Guwahati Hackathon | 2nd Runner-Up |
| Open Source Contributions | Contributions merged into Dify, Qdrant ecosystem, and CloudCV/EvalAI |
- Systems should scale, not just work
- Metrics drive decisions, not assumptions
- Latency and reliability are first-class concerns
Open to discussions around:
- AI systems engineering
- LLM infrastructure
- scalable backend systems