devwithjevClassifies and Routes Questions
Sid Sanghavi demonstrates using Jev to classify questions along known paths and route queries to a traditional LLM when appropriate. The demo compares Jev with Haiku on 100 questions.
View on X costfor 2x 100-query run, Jev was $0.006 and Haiku was $0.019
Sid Sanghavi@SiddhitSanghavi𝕏
Here's a simple demo of why and when you'd use #Jev from @typesafeai.
It doesn't generate text. It doesn't mean it can't hallucinate but if you know you have certain classification paths, it can classify with lightning speed and low cost and onward a query to the step where a traditional LLM is better.
Here, I do a walkthrough of a 100 questions with Jev versus Haiku and show why and when I'd use it. Not shown in video is the cost:
- for 2x 100-query run, Jev was $0.006 and Haiku was $0.019
- The second run was 1 question at a time, whereas the one shown is 4 questions at a time, but this
The displayed run processes four questions at a time; a second run processes one at a time. Sanghavi also reports the cost of two 100-query runs.
Also filed under Triage & routing
- Python Library for Jev Judgments
gut is an open-source Python library that turns Jev’s typed questions into a readable line of code.
- Real-Time Feedback Alert Classification
The author reports using Jev in a feedback platform, where real-time alert classification is already live.
- Yueli DEX Decision Exchange
Yueli DEX routes bounded business decisions to TypeSafe Jev or compatible providers and maps validated results to non-authorizing action intents.
- Medical Exam Option Scoring and Gating
Open Medical Jev is a frozen open-model medical decision stack that scores each exam option, then uses confidence and conformal gates to auto-release or escalate answers.