Dries Smit
AI Researcher at Tufa Labs
Zurich, Zurich, Switzerland
Summary
👤
Senior🎓
Top SchoolDries Smit is an AI researcher with eight years of experience specializing in reinforcement learning, multi-agent systems, and large language models, currently based in Zurich. He led the team that built InstaDeep's Laila, an LLM-powered lab assistant that integrates with real lab equipment, and contributed backend JAX code to the Mava MARL codebase. His work spans applied medical AI and foundational-model improvement, including reinforcement learning from execution feedback and inference-time compute optimization. Holding a PhD and MSc (cum laude) in Electrical and Electronic Engineering from Stellenbosch University, he blends rigorous academic training with hands-on engineering at research-driven companies and independent labs. An underappreciated strength is his ability to take production-oriented system design decisions in research codebases, bridging prototype R&D and deployable tooling.
8 years of coding experience
3 years of employment as a software developer
Doctor of Philosophy - PhD, Electrical and Electronics Engineering, Doctor of Philosophy - PhD, Electrical and Electronics Engineering at Stellenbosch University
Github Skills (6)
jax10
multi-agent-reinforcement-learning10
python10
computer-engineering7
machine-learning5
tensorflow4
Programming languages (4)
Github contributions (5)
instadeepai/Mava
Mar 2021 - Jan 2023
🦁 A research-friendly codebase for fast experimentation of multi-agent reinforcement learning in JAX
Role in this project:
Back-end Developer
Back-end DeveloperContributions:662 reviews, 1053 commits, 102 PRs in 1 year 10 months
Contributions summary:Dries's commits primarily involve refactoring and modifying files related to the "mava" codebase, which focuses on multi-agent reinforcement learning (MARL) in JAX. The changes include updates to training logic, executor code, and network structure to accommodate a specific network key feature. These changes were made across a variety of files related to MADDPG.
multi-agent-systemsmultiagentagentreinforcement-learningreinforcement-learning-agent
DriesSmit/CuriousAgents
May 2023 - Mar 2025
This repo explores ideas that can maybe in future be used to scale reinforcement learning to larger real world problems.
Contributions:8 PRs, 65 pushes, 6 branches in 1 year 10 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.