Jev CoLearn: Cafe Co-Learning Jev in HSR

​Description

​Jev launched three days ago. Nobody knows how to use it yet. Let's figure it out together.

​Jev is TypeSafe AI's first model, and it breaks the usual pattern: it doesn't write text. You ask it a question and it hands back a decision your code can act on directly. Pick one of many options, give a score, or answer yes or no, each with a confidence number attached. It's fast, it's cheap, and it's built for the thousands of small judgment calls buried inside real software: routing tickets, scoring leads, flagging risky actions, triaging documents.

​There are no tutorials, no best practices, and no public benchmarks. TypeSafe is asking developers to test it on their own use cases. So that's what we'll do.

​The plan

• ​A quick walkthrough of what Jev is and its three primitives: Choice, Score, and Noul

• ​Hands-on building: take a workflow you currently run through an LLM and see if Jev can do it faster and cheaper

• ​Build your own mini evals and compare results

• ​Share what worked, what broke, and what surprised you

​Who should come
If you are curious about Jev, then its for u.

​Before you come
Jev is currently in early access, so join the waitlist at typesafe.ai. No access yet? Come anyway and pair up with someone who has it. You can also access Jev through Open Router.

​Bring
Your laptop, a charger, and curosity.