ECE (AI) student at IGDTUW Delhi — building LLM evaluation frameworks, financial NLP pipelines, and research-driven AI systems. Agentic. Production-minded.
I'm a first-year ECE (AI) student at IGDTUW, Delhi with a focus on NLP, LLMs, and agentic systems. I build end-to-end AI pipelines — from LLM evaluation frameworks and sentiment classification to full-stack deployment.
Currently pursuing research opportunities in NLP and AI governance. Active as Tech Coordinator at IEEE IGDTUW, Chapter Lead at Youth Economy Lab, and Research Coordinator at E-Cell, where I author weekly analytical newsletters on emerging tech.
Intersection of AI systems and public policy — data protection, generative AI governance, and AI cybersecurity risk. Focused on how technical realities of LLMs (hallucination, bias, adversarial vulnerability) translate into regulatory frameworks, particularly in the South Asian context.
A growing stack across NLP, agentic AI, and full-stack development — grounded in strong fundamentals.
Technologies I work with
Real systems, real data, real impact. Each project is built end-to-end with research and production in mind.
Built an agentic evaluation framework benchmarking 5 LLM prompting strategies on 300 labeled financial sentences. Improved sentiment classification accuracy from 62% to 89.67% using Chain-of-Thought prompting — a +27% gain. Identified neutral bias and refined using anchor prompting with automated, reproducible pipelines.
End-to-end financial NLP pipeline using FinBERT for real-time news sentiment classification. Features a custom Sentiment Reliability Index (SRI — score 67.9) combining source credibility, recency, and model confidence. Integrated VIX volatility, price trends, and REST APIs; deployed on Streamlit.
Accessible community learning platform with Firebase authentication and real-time NoSQL database. Ranked Top 10 nationally (top 1.7%) out of 581 teams at HackwithMAIT 6.0. Accessibility-first, responsive design. Shipped concept-to-deployment in under 24 hours.
Open to research collaborations, internship opportunities, and conversations about NLP, AI governance, and LLM systems.