It may seem like something simple that you can do on your own, or even delegate to a junior member of your team. However, ML testing is much more complex when dealing with performance critical data products that require not just a solid model but one that can remain reliable for the foreseeable future. Also, many datasets today involve some PII, which needs to be handled properly, as we’ve stressed in previous episodes. What if you could do all that in a straight-forward, no-code, platform? Mr. Yunus Bulut, a co-author of mine on two occasions, joins me in this deep dive on the topic. He also covers the key aspects of his latest project, validAItor, an open-source platform for ML testing, to improve this whole process.
Podchaser is the ultimate destination for podcast data, search, and discovery. Learn More