Suggestions
Michael Troute
Senior Software Engineer @ Basis
Michael Troute is a Senior Software Engineer at Basis, located in San Francisco, California. He holds a degree in Computer Science from Stanford University, which provides him with a strong academic foundation in software engineering principles and practices.
In his current role at Basis, Troute is involved in developing and implementing software solutions, leveraging his expertise to contribute to the company's technology initiatives. His professional background includes various experiences that enhance his skill set in the tech industry, although specific details about his previous roles are not extensively documented in the available sources.
On LinkedIn, Troute is noted for having over 500 connections, indicating an active professional network within the tech community. His profile suggests he engages with technical topics and shares insights related to software engineering.12
Highlights
After 3 great years at Plaid, I’m excited to start a new adventure at Vanta!
In my current role, I don’t (yet) get to apply my knowledge about AI to my dat to day work. In retrospect, a concentration in Systems may have been more immediately relevant than one in Artificial Intelligence... However, a recent discussion about how our system could use AI to optimize usage of limited resources has me thinking about old school work.
I dug up this paper in which two brilliant friends (Brahm Capoor, Ali Malik) and I used deep learning to generate GIFs from individual “seed” frames. Somehow we thought it would be funny to use the most complex model we knew how to construct to do something as dumb as making more GIFs (as if the internet needed more GIFs).
The resulting model was (1) recurrent, (2) adversarial, (3) generative, (4) convolutional, (5) auto-encoding, and (6) ridiculous. The images it produced were very strange and weirdly lifelike. We built the model in the hopes that it would be able to predict and generate the trajectory of entities moving in straight lines and “bouncing” realistically when they hit the edges of the frame. Instead, the videos generated looked like cells under a microscope—moving around organically and even (though of course this was an accident) producing “offspring” in some cases.
Attached, for your reading pleasure, if you’re into this sort of thing.
