Jim D'souza
ML Tech Lead - Behaviors at Woven by Toyota
San Francisco, California, United States
Summary
👤
Senior🎓
Top SchoolJim D'souza is an experienced robotics and autonomy leader with 13 years building perception, prediction, and planning systems for self-driving vehicles and spatial AI. Currently ML Tech Lead for Behaviors at Woven by Toyota, he has led global teams and set technical direction for large, cross-border ML organizations focused on end-to-end autonomous driving. Previously he scaled ML-behaviors at Torc Robotics, delivering learned pixel-to-plan models and transitioning fleets from analytical to ML-native planners for heavy vehicles. Jim combines deep hands-on expertise in sensor fusion, SLAM, motion forecasting and real-time embedded systems with a track record of shipping vehicle demonstrations and peer-reviewed research at top conferences. Comfortable across research and production, he repeatedly bridges academia and industry—having collaborated with Princeton and Mila—while keeping a playful practitioner’s focus: he genuinely likes robots that move. Based in San Francisco, he brings cross-cultural leadership across US, Europe and Japan to solve autonomy’s hardest integration challenges.
13 years of coding experience
9 years of employment as a software developer
Bachelor of Technology - BTech Electrical and Electronics Engineering, Bachelor of Technology - BTech Electrical and Electronics Engineering at National Institute of Technology KarnatakaTechnical University of DenmarkMaster of Science - MS Robotics, Master of Science - MS Robotics at Nanyang Technological University Singapore
Stackoverflow
327reputation
13kreached
8answers
8questions
Github Skills (174)
px410
python10
unbound10
wireguard10
pihole10
deep-learning10
apollo10
ros10
autopilot10
mavlink10
adblock10
ml10
neural-network10
drones10
tensorflow10
Programming languages (14)
Github contributions (5)
jimdsouza/OnlineClassroom
Oct 2015 - Oct 2015
Contributions:29 commits, 18 pushes, 1 branch in 13 days
jimaldon/wirehole
Jan 2022 - Feb 2025
WireHole is a combination of WireGuard, Pi-hole, and Unbound in a docker-compose project with the intent of enabling users to quickly and easily create a personally managed full or split-tunnel WireGuard VPN with ad blocking capabilities thanks to Pi-hole, and DNS caching, additional privacy options, and upstream providers via Unbound.
Contributions:13 pushes, 4 branches in 3 years 1 month
composeupstreamsplit-tunnelholedns
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.