Engineers show Jev system-one models inside application code at AIE Code
Original title: The Jev Revolution: What Engineers Are Already Building with System One Models
Glean, UiPath, and Workday described ranking, retrieval, and UI prototypes that hide whether Jev or an LLM is used.
At the AI Engineer Code Summit, TypeSafe AI CEO Diogo Almeida argued that Jev, a non-generative model released in September, should be called from application code for faster, cheaper structured decisions. He called it reliable intelligence and urged engineers to automate more with it. Glean said Jev handles changing taxonomies without repeated fine-tuning, while UiPath and Workday showed ranking and interface prototypes.
Unconfirmed: Usage claims come from vendor and customer presentations, not independent measurements.
- Almeida said LLMs can be wrong about 5% of the time and still block automation.
- UiPath ranked document chunks with thousands of Jev calls in under a second.
- Glean said customers need not know whether a flow uses Jev or an LLM.
- Workday prototyped a time-off card that fills in while the user types.
- Vendors
- TypeSafe AI
- AI models
- Jev
From the original article
4 min read at latent.space
At AI Engineer Code Summit, lead Jev creator Diogo Almeida called on developers to automate everything. Meanwhile, Glean and others showed how they’re using Jev.
By Richard MacManus · latent.space
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.