# About Mohit Goyal

I build agentic AI systems, RAG infrastructure, and applied LLM research projects that move from paper-level ideas into usable products. Most of my work lives at the edge of research and shipping: reliable systems, not just impressive demo windows.

Mohit is currently splitting his time across three overlapping roles: building the candidate-discovery layer of an internal AI recruiting platform at Lenskart, shipping production agentic AI features (artifacts, agent skills, multi-agent orchestration, and RAG pipelines) at C3alabs, and working full-stack across ArmorIQ, from an SDK observability system and a documentation MCP server to security and performance hardening.

Before that, he worked as an ML researcher at RAAPID INC, co-developing GRIT, a geometry-aware parameter-efficient fine-tuning method that combines K-FAC preconditioning, Fisher-guided reprojection, and dynamic rank adaptation to fine-tune under 1% of a model's parameters while outperforming full fine-tuning and LoRA on standard benchmarks. He has also built autonomous agents at Kartavya Technology (market research, voice interviewing, credit risk scoring, outbound calling) and has three publications spanning parameter-efficient fine-tuning, zero-trust security with explainable AI, and a peer-reviewed review of Microsoft Semantic Kernel.

Across all of this, his focus areas are Agentic AI, RAG infrastructure, Applied LLM research, MCP systems, worked mostly in Python, PyTorch, FastAPI, MCP, LangGraph, Next.js. He has also won 15 national-level hackathons, reflecting a bias toward shipping working systems quickly under constraints, not just writing papers.

He is based in Gurugram, India, and is reachable via the contact page or any of the social/profile links below for collaboration, research discussion, or roles in agentic AI, ML engineering, or full-stack AI product work.

## Current roles

- **AI Engineer Intern** at [Lenskart](https://www.lenskart.com/), Gurugram, India (On-site) — July 2026 to Present.
- **Software Engineering Intern** at [C3alabs](https://www.c3alabs.com/), United States (Remote) — October 2025 to Present.
- **Full Stack Developer Intern** at [ArmorIQ](https://armoriq.ai/), Remote — June 2026 to Present.

## Publications

- **GRIT: Geometry-Aware PEFT with K-FAC Preconditioning, Fisher-Guided Reprojection, and Dynamic Rank Adaptation** (arXiv preprint, 2026) — [arXiv](https://arxiv.org/abs/2601.00231)
- **AoT-BTS: An Explainable Age-of-Trust Based Behavioral Scoring Framework for Real-Time Zero Trust Enforcement using SDN** (Research Square, 2026) — [link](https://doi.org/10.21203/rs.3.rs-9722330/v1)
- **A Comprehensive Review of Microsoft Semantic Kernel: Architecture, Capabilities, and Research Directions** (IEEE DELCON, 2025) — [link](https://doi.org/10.1109/DELCON68055.2025.11400282)

## Focus areas

- Agentic AI
- RAG infrastructure
- Applied LLM research
- MCP systems

## Stack

- Python
- PyTorch
- FastAPI
- MCP
- LangGraph
- Next.js

See [/work-experience](/work-experience) for the full role history and [/projects](/projects) for shipped work.