Radar Rising Jev Use casesCases MethodSubmit

Close

×

Radar

Missed a build?

Paste an X post or GitHub repo we missed. It is reviewed first, then scored by the normal pipeline.

X or GitHub URL

𝕏Built by @psingh

Loading

Independent radar of Jev builds, experiments and use cases — scored by Jev, not by likes. Not affiliated with TypeSafe AI.

Built by @psingh

Explore

Jev Radar
Rising
How Jev works
Jev use cases
How Jev Radar works

Use cases

Routing
Classification
Agents
Search & Retrieval
Developer Tools
Automation

Elsewhere

TypeSafe AI
Follow me on X

1857 published builds · 7098 buildersBuilt with Amp · © 2026 Jev Radar

CHOOSEData & productivity

Police Crash Narrative Conversion to Probabilistic Variables

Natural Language Processing Papers@HEI

Crash datasets that carry an investigator narrative hold information the coded fields omit. Coding those narratives at scale has been blocked by three obstacles. Frontier large language models are costly at that scale, their generated text cannot be verified, and no rule says how much output a human must check. This paper formulates narrative coding as gated, typed decisions answered by Jev, a System One model that returns probabilities over analyst-defined options and generates no text. A screen covered 499,500 Texas narratives and 195,857 were coded with a 27-question schema. Cost is governed by schema size rather than narrative length. The probabilities are audited against coded fields and against 2,416 blinded human judgments drawn under a stated sampling design. Two frontier large language models are benchmarked on the same records. Against human labels the typed model attains an F1 of 0.908. One frontier model gains 0.059 and the other is indistinguishable from it. Calibration va

Sep 24

38 views on X · does not affect score

This project converts police crash narratives into probabilistic crash variables using a System One Model, referred to as Jev.

Calibrated Decisions at Scale: Converting Police Crash Narratives into Probabilistic Crash Variables with a System One Model (Jev) Amir Rafe, Subasish Das arxiv.org/abs/2609.24052 [𝚌𝚜.𝙲𝙻] x.com/HEI/status/2103052792147456510/photo

Use cases

Classification

Related builds

Jev Model Asymmetry Analysis

This project analyzes the Jev model's behavior, revealing that natural language wording induces a significant conviction asymmetry between buy and sell sides, while the model itself is symmetric when presented with symmetric prompts.

Dan Holzrichter@dholzric

83 views on XJev score 79

TypeSafe Computer Use

This project enables a Mac to achieve plain-English goals by combining OCR and the Accessibility tree to drive actions, avoiding large model loops for screenshots.

Yash Chaudhary@YashChaudhary

4 views on XJev score 78

Jev Decision Model Evaluation

Jev, a decision model from TypeSafe, was tested on standard AI problems, answering structured questions with probabilities and performing comparably to or better than other models on single-step tasks, but struggled with sequential reasoning without state information.

Nikhil Deshpande@nikhdesh

113 views on XJev score 76

Project photo 1

PostgreSQL Extension for TypeSafe AI Jev Integration

This project is a PostgreSQL extension that allows SQL queries to directly call TypeSafe AI's Jev for classifying rows within a database table, avoiding the need to export data for external processing.

giulio@giuliosmall

4.4K views on XJev score 76

Quality78

NoveltyiHow surprising the Jev use is versus an obvious typed-decision call. Generic mentions score low; a new primitive scores high.73
UsefuliHow reusable the technique is for another builder. Toys score low; a production-quality idea scores high.83
SubstanceiHow much concrete evidence supports the claim: code, numbers, demo, architecture. Hype scores low.82

Quality is Jev’s judgment of the idea. Likes, reposts and views do not affect it. Relevance is a separate gate.