- 🎓 MTech in Data Science, IIT Guwahati (CPI: 9.61) | BTech in Electrical Engineering, VJTI Mumbai
- 💼 ML Engineer / Data Scientist with ~2.5 years of ML/AI experience — NLP pipelines, LLM-based RAG systems, and cloud-scale analytics on Azure & GCP
- 🔬 ACM Published — S-VQA at ICVGIP 2023
- 🏆 LeetCode Knight — Top ~5% globally
- 🚀 Passionate about LLMs, RAG, PEFT/LoRA fine-tuning, Agents, and NLP evaluation
- 🌱 Currently exploring: systematic ML evaluation, long-document QA, and agent orchestration patterns
- Core areas: LLMs · RAG · PEFT/LoRA · NLP · Computer Vision · Transformers · ML Evaluation
- Cloud: Azure (Databricks, AI Search, AI Foundry) · GCP (BigQuery, GCS)
- Languages: Python · SQL · C/C++
| Project | Description | Links |
|---|---|---|
| Diagnosing RAG Failure Modes | Systematic evaluation of RAG on long-document QA (QASPER dataset). 4 stress-test experiments, custom Evidence Coverage Score (ECS) metric. | Live Demo · GitHub |
| S-VQA | Sentence-based Visual Question Answering — TDIUC-SVQA dataset construction and multi-task multimodal modeling. ACM · ICVGIP 2023. | GitHub · Paper |
| Business Analytics Chatbot | Conversational analytics agent built with Google ADK + Gemini. Natural language queries returning concise answers and charts. | GitHub |
| PEFT LoRA Fine-tuning | Parameter-efficient fine-tuning of LLMs using LoRA — with a live interactive demo on HuggingFace Spaces. | Live Demo · GitHub |
| Movie Recommendation System | Collaborative filtering on the MovieLens dataset with recommendation-specific evaluation including Long Tail analysis. | GitHub |
| Article Bias Prediction (LSTM) | Multi-approach similarity-based political bias detection in news articles using LSTM. | GitHub |
| Image Caption Generator | Automatic image captioning model inspired by the Show and Tell architecture. | GitHub |
S-VQA: Structured Visual Question Answering
ACM · ICVGIP 2023
📖 View Paper → | 💻 View Paper Summary →
loT Based Real-Time Harmonic Monitoring System for Distributed Generation
IEEE · I2CT 2018
📖 View Paper →
- 🔍 Open to MLE / Applied Scientist / Senior Data Scientist roles at product-based companies
- 🛠️ Building: systematic RAG evaluation frameworks and agentic AI workflows
- 📬 Best way to reach me: LinkedIn
