TLDRocket
Sign in

Workday is using TypeSafe’s Jev to build user interfaces on the fly

The New Stack Frederic Lardinois

Workday moved Sana to the front of its HR, finance and IT tools. It’s using TypeSafe’s Jev to build the screen live, not from a fixed app.

Based on reporting by The New Stack, Frederic Lardinois — read the original for the full story.

Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error

Workday put Sana, its AI assistant, at the center of its HR, finance and IT software on Tuesday. That sounds like a routine product update until you look at how the interface is assembled: Workday says Jev, TypeSafe’s decision model, is choosing the pieces of the screen as the session unfolds.

In the demo, that meant a question about a paycheck brought up a card with the amount hidden until a click, while a request for time off produced a date picker and Sana filled in the rest. Joel Hellermark, Workday’s chief AI officer and Sana’s founder, said the usual way of generating an entire interface is slow. His pitch is simpler: give the model the design components, let it pick, and the UI appears almost instantly.

Workday says it handed Jev every atomic component in its design system and didn’t hard-code the selection logic. The catch was data preloading, which the team still had to get right. In the demo, the paycheck card appeared before Hellermark had even finished typing.

TypeSafe says Jev responds in 70 to 500 milliseconds, which is exactly the kind of speed this sort of interface needs. The model doesn’t write text. It chooses from up to 255 options, or gives a score, or says yes or no, each with a probability attached.

There’s a lot of momentum around this approach. TypeSafe launched Jev on Sept. 15 with a $40 million seed round led by DCVC, then on Friday said it had raised an $870 million Series A led by Andreessen Horowitz at a $7.5 billion valuation. Workday has already added TypeSafe to its list of subprocessors, and TypeSafe says its hosted API runs in the U.S. with zero data retention for enterprise customers. But the important detail is still the product one: Workday is betting that the best AI interface may be the one that never fully gets written in the first place.

My take — AI-written commentary, not fact-checked reporting

This is the sane version of AI UI: pick from known components, don’t let a model freestyle the whole thing and call it product design. The industry keeps pretending every problem needs a chat window; sometimes it just needs a fast chooser with a short leash. Turns out the cage is doing a lot of useful work.

Read more about this at: The New Stack

Related stories

The daily briefing

Every AI story that matters, in your inbox by 8am.

TLDRocket reads all relevant sources, removes duplicate coverage, and summarises the day in two minutes. Follow companies and topics for alerts, or get the briefing in Slack. Free, no spam, unsubscribe anytime.