I recently cofounded a startup with the goal of automatically labelling defects in sewer inspection videos. Yes, I really did say “sewer”. It’s actually a huge industry($50B) and the crumbling sewer infrastructure is a real problem. Right now around the country sewer inspectors are driving small tethered cameras around the sewers and every time they see a crack or a root or grease they stop driving the camera and write down what they see into a database. They are watching a combined 15,000 hours of video footage EVERY day. It’s insane. It’s slow, expensive and inconsistent.
So we are building an AI tool to help them do it faster with the ultimate goal of automatically watching the videos and labelling the defects automatically. My cofounder and I have been working in the industry for many years.
I’m a huge fan of fast.ai. I’ve taken many of the courses including part 1 2019 in person in SF.
We are looking for another engineer to join our team. Ideally you have lots of python and fast.ai experience. Right now we are concentrating on more deep learning vision related prediction tasks so looking more for experience on that end.
Please send an email to jobs@sewerai.com if you are interested.
Thanks
-Matt