Manav Pandey
I build production AI systems — from agentic frameworks serving thousands of developers to fine-tuned open-weight models running in production. I’m drawn to the boundary where engineering meets research, especially energy-based approaches to reasoning.
Currently a Senior AI Research Engineer at American Express and an M.S. student at Georgia Tech. My side research explores whether energy-based models can offer a principled alternative to autoregressive reasoning — Enso is my first proof of concept.
If I had to describe myself
01Production Python is my foundation — every system I’ve shipped started here, from conversational AI to agentic frameworks to internal developer tools.
Python Engineer.
02I’ve fine-tuned, quantized, and deployed open-weight models in production — the kind of work where inference latency and model quality both matter.
ML Engineer.
At Lightsource I was the sole ML engineer: vLLM serving across multiple GPUs, dynamic LoRA adapters, DPO and PPO fine-tuning of Mistral and Mixtral, INT4-FP8 quantization for production inference. At Amex I built the internal model routing and tooling layer that sits between developers and LLMs. The craft is making models work reliably at scale, not just on a notebook.
03I’m pursuing a research question: can energy-based models learn to reason through optimization rather than token prediction?
Researcher.
Enso is my first attempt at an answer — a 36M-parameter model that solves Sudoku via Langevin dynamics in a learned energy landscape. I reproduce results before I trust them; Enso started as a replication of Kona 1.0 and became something new. I’m now studying ML formally at Georgia Tech while continuing to build in this direction.
Experience
Senior AI Research Engineer
American Express — Enterprise Data & AI R&D
- Built and shipped American Express’s first agentic use case in production (PR Agent) using LangGraph and MCP.
- Lead engineer on internal GenAI frameworks serving 4,500+ developers — building Amex-native equivalents of LangChain, LangGraph, OpenAI-compatible model routing, and MCP tooling.
- Core GenAI educator at Amex, leading technical sessions distilling cutting-edge research (MLA, Muon Optimizer, Context Engineering) for 500+ engineers.
- Python Guild AI Lead — drove best-practices architectural decision records for ML/AI and merged the official Amex Python Skill into the internal AI skills repository.
- Filed 4 patents with 28 additional filings underway across self-supervised learning and agentic AI.
Machine Learning Engineer — Research
Lightsource
- Sole ML Engineer — deployed open-weight GenAI models into production using vLLM across multiple GPUs with dynamic LoRAs.
- Fine-tuned Mistral 7B and Mixtral 8x7B for multilingual content generation using DPO and PPO with synthetic preference data.
- Optimized inference with INT4-FP8 quantization; implemented spherical interpolation model merging for improved multilingual performance.
Director of AI
Curiouser
- Led engineering for a conversational AI platform. Built LoRA dynamic adapter selection, chain-of-thought tool calling, and autonomous knowledge graph creation for persistent user understanding.
Software Engineer
American Express
- Built production conversational AI using BERT models achieving 90% accuracy across 1M+ monthly interactions.
- Implemented APM monitoring across 100+ microservices using Splunk and OpenTelemetry.
ML Research Assistant
Texas A&M University
- Enhanced SVM accuracy from 79% to 94% through kernel optimization for sentiment classification and cross-border threat detection.
Projects
Energy-Based Model for Constraint Satisfaction
A personal research project: a 36.5M-parameter model combining JEPA-style joint embedding with energy-based inference, solving hard Sudoku through Langevin dynamics in latent space — exploring whether energy-based optimization can substitute for autoregressive generation.
- 96.6% puzzle accuracy — exceeding Kona 1.0’s open-source benchmark of 96.2%
- Forward pass achieves 95.6%; Langevin dynamics adds +1.0% through test-time compute scaling
- Uses mechanistic interpretability to analyze energy-based model reasoning
Dialogue Tree Search
MCTS-Inspired Synthetic RL Data Generation
A parallel beam search system that treats conversation trajectories as a search tree, using Monte Carlo rollouts to explore diverse dialogue paths. Generates synthetic preference datasets for training tool-using agents via GRPO and PPO with Elo-based scoring.
- MCTS-inspired parallel beam search over conversation trajectories
- Produces preference data for RL fine-tuning of tool-using agents
Education
Georgia Institute of Technology
M.S. Computer Science (Machine Learning)
2025 – Present
Texas A&M University
B.S. Computing
2018 – 2022
Awards & Recognition
American Express Inventor Award
2025
Recognized for 4 filed patents (28 additional underway) spanning agentic AI, applied ML infrastructure, and self-supervised learning.
Anthropic Bug Bounty Program
Contributed to AI safety through Anthropic’s bug bounty program, identifying vulnerabilities in foundation model behavior.