Akash Velu
Summary
🤩
RockstarAkash Velu is an engineer focused on AI, robotics, and machine learning with eight years of experience bridging academic research and production systems. He holds EECS degrees from UC Berkeley and an MS in Computer Science from Stanford, where he contributed to robotics and reinforcement learning research (see arXiv:2307.11897) and supported core AI courses as a course assistant. His industry experience spans ML roles at Tesla and Waymo and internships at Google and Numenta, delivering end-to-end deep learning for autonomy and neuroscience-inspired ML research. At The Bot Company he now builds robots that save people time, drawing on hands-on systems work from simulation to scalable behavior prediction. An active open-source contributor, he improved documentation and code quality for the MAPPO multi-agent reinforcement learning implementation, highlighting his emphasis on maintainable, readable ML code. He pairs a research-first mindset with practical engineering discipline, comfortable moving ideas from papers into production.
9 years of coding experience
Github Skills (7)
algorithms10
pytorch10
multi-agent10
ppp10
python10
documentation9
machine-learning9
Programming languages (3)
Github contributions (5)
marlbenchmark/on-policy
Feb 2021 - Jul 2021
This is the official implementation of Multi-Agent PPO (MAPPO).
Role in this project:
ML Engineer
ML EngineerContributions:38 commits, 7 pushes, 2 comments in 4 months
Contributions summary:Akash primarily focused on improving the documentation and code structure within the multi-agent reinforcement learning framework. Their commits involve adding detailed docstrings and descriptions for classes and methods within the core actor-critic architectures. Additionally, the user removed unnecessary imports and performed style fixes to improve code readability and maintainability. The user's work directly supports the development and understanding of the MAPPO algorithm within this repository.
multi-agenthanabimapposmacmpes
akashvelu/nupic.research
Jan 2021 - Jun 2021
Experimental algorithms. Unsupported.
Contributions:51 pushes, 4 branches in 4 months
data-structuresunsupported
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.