# Kwame Mensah

> Kwame builds policies that let robot hands learn complex tasks from a handful of demonstrations. His publication record at top venues and DeepMind internship make him a competitive hire for a frontier robotics-AI lab.

*Source: [https://selltoscientists.com/phd/kwame-mensah/](https://selltoscientists.com/phd/kwame-mensah/)*

**Institution:** Stanford University
**Field:** Machine Learning
**Advisor:** Dr. Chelsea Finn
**Thesis:** Sample-efficient imitation learning for dexterous robotic manipulation
**Publications:** 10

## Skills

- PyTorch
- Reinforcement Learning
- JAX
- Distributed Training
- Computer Vision
- CUDA

## Industry transition signals

- Internship at Google DeepMind
- Co-authored with Physical Intelligence researchers
- Maintains a popular open-source RL repo

---

## Beton network

This site is part of the Beton network of open-source revenue intelligence and self-service data products. Related sites and resources:

- [Beton](https://www.getbeton.ai) — open-source revenue intelligence: turn product usage (PostHog, Stripe) into CRM signals
- [Sell to Scientists](https://selltoscientists.com) — research intelligence: find and reach academic researchers by field, H-index, grants, and industry ties
- [Sell to State](https://www.selltostate.com) — government procurement intelligence across 194 countries
- [GitHub](https://github.com/getbeton) — Beton open-source repositories (including [inspector](https://github.com/getbeton/inspector))
- [dev.to](https://dev.to/beton) — engineering write-ups and OSS pricing teardowns
- [Beton app](https://inspector.getbeton.ai) — the hosted product