What advice would you give to CTOs newly managing machine learning (ML) and data science teams?

Talent Management & Performance IT Strategy & Roadmap Data Science & Algorithms

2.2k views4 Comments

Sort By:

Oldest

CTO in Services (non-Government)2 years ago

If you've not dealt with R&D and all of a sudden you have one of these teams under you, allow the staff to govern the direction they’re heading. Your job as a CTO is to give them the general direction their work's meant to take. You’re only going to make your staff leave by insisting on delivery dates and written weekly updates, or subscribing to a ticket story system.

If you make them leave you might lose a valuable R&D researcher. And for every 10 applicants that I get on paper, nine of them are not a good fit. On paper, they've got the right academic background, qualifications and sometimes even a bit of industry experience, but only 1 in 10 will be a valuable addition to the company. So if you keep cycling though your staff, you'll soon be left with the people that are not that valuable to your company. And those people are not going to leave because they're not the ones that are being headhunted.

CEO, MSSP - High Assurance Cybersecurity SOC in Services (non-Government)2 years ago

Give them the latitude to experiment, fail, and rebuild... that the is premise of ML and AI. But most importantly, keep the objective and end-state in focus while innovating.

Director of IT in Software2 years ago

Pay particular attention to the data formate/readiness. If your data is not clean/ready for ML, your results will not be valuable. They say Garbage in, garbage out.

Please join or sign in to view more content.

By joining the Peer Community, you'll get:

Peer Discussions and Polls
One-Minute Insights
Connect with like-minded individuals

CIO in Services (non-Government)2 years ago

Put together a lab for them to build and break everything; the more they can play with it, and the more they can experiment, the better it is for their understanding of the technical side of the project. I'm a firm believer that all new projects should be implemented way ahead of roll-out, by putting it all together in a learning lab environment. I'm a "learn it by doing it and breaking it" kind of guy, and a lot of my engineers and techs are the same.