Starting from the person, not the dataset
What changes about a project when the outcome is somebody's day rather than a metric.
Speaking topic
The work behind the TED talk. Where machine learning is pointed at problems that are about people rather than throughput, what that changes about how you build, and what it demands of the people building it.
The room
Healthcare and public sector audiences, purpose-led conferences, universities, and any programme with a social impact track.
Every booking starts the same way: tell him who is in the room and what you want them doing afterwards. The enquiry form asks for exactly that and nothing else.
What it covers
Shaped to your audience rather than delivered off the shelf, so treat these as the argument rather than as a running order.
Starting from the person, not the dataset
What changes about a project when the outcome is somebody's day rather than a metric.
Accessibility as an engineering problem
Work on communication and access, and what it taught him about designing for people the default case ignores.
The responsibility that comes with it
Where the stakes are highest and where the temptation to overclaim is strongest.
Why him
This is the subject of his TED@IBM talk and of his TEDxCincinnati talk, both public and both linked on this site.
On this subject
Every one is public, plays here, and carries its year. This is an archive, not a schedule.
Questions
He speaks to the projects that are public and documented. He does not present private or clinical work, and he will tell you in advance exactly which examples the talk uses.
Yes. It is the talk that travels furthest outside technical rooms, and it is the one most often programmed as an opening or closing keynote.
Bring Tanmay to your stage.
Send the date and tell him who is in the room. Availability comes back first.