evaluate( )





Pattern
Jev Semantic Clustering
Discover themes in comments with no labels up front. Embeddings pick likely pairs, Jev answers one boolean per pair, and code does the grouping.
IntermediateOct 2026View the full pattern on desktop


Problems solved
- Find themes in feedback without predefined categories
- Keep grouping deterministic while a model makes the semantic calls
- Show what Jev adds over clustering embeddings alone
Use cases
- Audience themes from video or post comments
- Grouping support tickets or survey answers
- Discovering categories before writing a classifier
AI SDK APIs
experimental_evaluateembedManygenerateTextgateway
Tags
jevevaluationtyped-decisionsbooleanclusteringembeddingsai-sdk
Requirements
- Node.js
- 24+
- Memory
- 256MB
- Other
- Vercel AI Gateway API key
Environment variables
AI_GATEWAY_API_KEYGet key
External services
vercel-ai-gatewayupstash
Ready to use?
View the full interactive pattern with live preview and copy the source code.
View the pattern on desktopRelated patterns


Jev Output Verifier
Write an answer, then show it only if Jev is at least 90% sure it actually answers the question. If not, rewrite once.
jevevaluation+4
View pattern →


Jev Loop Breaker
Look up facts one step at a time. Stop when Jev is at least 90% sure the notes answer the task, or after 3 lookups.
jevevaluation+4
View pattern →



