AI Engineer with 6+ years in Agentic AI Systems & LLM Toolchains
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
AI/ML Engineer and M.S. Computer Engineering student at NYU with 3+ years of production experience building agentic AI systems, multi-agent workflows, LLM toolchains, and MCP-integrated pipelines. Fluent in responsible AI devel-opment, adversarial prompt evaluation, and first-principles decomposition of complex engineering problems. Passionate about applying agentic AI to accelerate technical workflows from code generation to automated reasoning pipelines with a deep commitment to interpretable, trustworthy AI.
New York University (Tandon School of Engineering)
Master of Science · Computer Engineering
September 1, 2025 – May 1, 2027
Dr. A.P.J Abdul Kalam Technology University
B.Tech · Computer Science and Engineering
August 1, 2017 – August 1, 2021
New York University
Teaching Assistant
January 1, 2025 – Present
India
Coforge
Senior AI Engineer
September 1, 2024 – August 1, 2025
India
LetBrief
Co-Founder
September 1, 2024 – Present
India
mavQ
Associate Machine Learning Engineer
February 1, 2022 – September 1, 2024
India
Geo-Insights-Pro: Multi-Agent Orchestration Platform
June 19, 2026 – Present
Built a multi-agent system with explicit task decomposition and sub-agent handoffs – agents independently call external tools (Tavily web retrieval, Gemini reasoning, PostgreSQL storage) via MCP-style tool-use patterns. Implemented responsible AI checks: hallucination detection, brand-accuracy auditing, and structured evidence tracing – ensuring all LLM outputs are grounded and auditable before surfacing to end users.
Reinforced Weak-to-Strong Teaching: Agentic Code Generation via GRPO
June 19, 2026 – Present
Designed a sub-agent reasoning framework where a teacher agent generates structured hints to guide a student LLM toward correct code solutions - using Group Relative Policy Optimization (GRPO) as the reward signal. Achieved 72.5% HumanEval pass@1 (2.2× baseline) and 59.6% MBPP, outperforming SFT by 11% - demonstrating that adversarial hint quality and reward-signal design are the primary levers in agentic code generation.
Gmail RAG Agent
June 19, 2026 – Present
Deployed a fully local agentic retrieval system over Gmail history using pgvector + Ollama – no cloud dependency, demonstrating sovereign, edge-ready LLM pipeline design.
AWS Certified Cloud Practitioner
AWS
June 1, 2026 – Present
Prompt Engineering for ChatGPT
Unknown
June 1, 2026 – Present
Advanced SQL
Unknown
June 1, 2026 – Present
Python Data Structures
Unknown
June 1, 2026 – Present
EDA for Machine Learning
Unknown
June 1, 2026 – Present
Cultural Fit Analysis
The candidate's diverse project portfolio, including personal projects and startup experience, demonstrates initiative and a broad interest in AI applications. Their involvement in a startup (LetBrief) and academic role (NYU TA) shows adaptability and a willingness to take on varied responsibilities. The focus on interpretable and trustworthy AI aligns with best practices and a responsible development culture. The breadth of skills across AI/ML, MLOps, and data engineering indicates a versatile team player.
Soft Skills & Operational Fit
The candidate's experience as a Co-Founder and Teaching Assistant suggests strong leadership, problem-solving, and communication skills. Their work on responsible AI checks indicates a methodical and ethical approach to development. The emphasis on 'first-principles reasoning and systematic debugging' aligns with a robust operational fit for complex problem-solving.