About
What he does now, and the argument he has been making since the first time he stood on a conference stage.
Tanmay Bakshi is an AI and software architect at IBM, working on Watson applications, and a Google Developer Expert for machine learning. He has written four books, for Manning, Packt and McGraw Hill, and his YouTube channel has 324,000 subscribers.
He first argued on the nasscom stage in 2017 that AI raises the ceiling on what people can do rather than replacing them. He returned to India in 2026 to open INFOCOM Hyderabad and to speak at People Summit in Bengaluru, and the argument had not changed. What changed is that there is now nine years of evidence for it.
His current work is about the part nobody had to think about in 2017: what happens when the systems act on their own, and who is accountable when one of them is wrong.

IBM
AI and software architect, Watson applications.
Developer Expert for machine learning.
TED
TED@IBM, and TEDxCincinnati.
Four books
Manning, Packt, and McGraw Hill.
324,000
Subscribers on his YouTube channel.
The record
Dated, in order, including the stretch where there is nothing to show. He was building rather than touring.
Technology that tackles the teen suicide epidemic. The first of the talks that are still the ones people find.
A Linux Foundation main stage keynote on where the tools were heading, and a TEDx talk on connecting the disconnected.
The opening keynote, and the first time the argument he still makes was the whole talk: AI raises the ceiling rather than replacing the person.
Hello Swift for Manning, Cognitive Computing with IBM Watson for Packt with Rob High, and the two McGraw Hill titles on Julia and Go. The teaching channel passed three hundred thousand subscribers.
Almost no public speaking. He was building rather than touring, as an AI and software architect at IBM. Any page that shows a continuous run of appearances through here is smoothing over a real gap.
Governing the second workforce: autonomous agents, and who is accountable when one of them is wrong. The subject that is now most of his work.
An opening keynote and a masterclass session, nine years after the nasscom stage, making a version of the same argument with nine years of evidence behind it.
What he argues
AI that augments
The argument Tanmay has been making for nine years, and the one he still opens with: the useful question about AI is not which jobs it removes, it is which ceilings it lifts.
The second workforce
Autonomous agents are being handed real work with real consequences, and most organisations have no answer to the question of who owns a decision an agent made alone.
Signal, not hype
A talk for the people who have to make decisions with money attached.
AI for human problems
The work behind the TED talk.
Bring Tanmay to your stage.
Send the date and tell him who is in the room. Availability comes back first.