Jev Cookbook

AIEdTalks part 4 on the batching trick: a loop that resends a long document for every question, versus one call that asks all thirteen.

The index of TypeSafe's worked examples, from parallel questions and reranking to guardrails, date extraction, and self-consistency, each with datasets and measured results.

A Japanese demo that asks Jev 0.5 seconds after speech ends whether the turn is really over, then extends the hold time based on the returned score.

Semantic search over GitHub's Terms of Service: one request ranks all 218 lines with a Choice while a Noul checks whether the document contains an answer at all, including when it should say no.

A two-stage extraction cascade: a mini model extracts, a Noul battery verifies each field in one request, and a 0.7 gate escalates to a reasoning model, sitting on the cost/quality frontier.

Turn natural language into typed function calls: closed-set arguments become Choice questions, optional ones become Noul questions, and 54 questions ship in one request for a trading assistant.

Decide which of 450 candidate product pairs from two beer catalogues describe the same entity: one three-level Score per pair splits 40 merges, 50 curator reviews, and 360 non-matches.

Catch wrong and hallucinated citations against RFC 7519 with a string match plus one Choice question: four accurate citations verified at 0.93 or more and all four planted failures caught.

A repeatability study: an 8-question moderation rubric run 15 times per condition shows a mean probability standard deviation of 0.0098, and a 0.60 uncertainty gate lifts agreement to 99.2%.

A 14-Noul claims-triage rubric over one insurance claim, repeated 15 times: mean probability standard deviation of 0.0102 at 111 ms per call, with a 0.30 to 0.70 band for human review.

Extract dates by asking seven Choice questions about a date's shape and parts, then resolving them in code: five of six examples auto-accepted and missing dates flagged at 0.46 confidence.

Score retrieved passages with four Noul questions each, then route in code: prompt injections dropped, false premises kept as conflicts, and evidence assembled separately for the generator.

Rank 182 agent skills in one request and re-read the top three in a second: over 488 requests, wrong skill loads fell from 16.8% to 7.3% and needless loads from 9.8% to 4.0%.

Reconstruct Markdown from unformatted text in two requests: 16 Noul line-pair questions stitch split sentences, then 62 questions classify each block's type and companions.

Turn wine tasting notes into 67 numeric columns for a CatBoost regressor: five rounds of reading its own errors cut held-out RMSE from 3.09 to 1.77 across 800 unseen reviews.

Regex finds candidate emails, phone numbers, and amounts; Jev picks the requested span so code copies a verbatim value that cannot be invented, with no free-text extraction required.

A 13-question regulatory briefing that shows batching every question into one call is 12.2x cheaper and 10x faster with no change in answers.

One question per query-candidate pair raises top-1 accuracy from 5% to 18% and top-10 accuracy from 38% to 62% on 40 CLERC legal queries.

Screen every message going into and out of an LLM app: a Noul describes possible hazards, a Score rates severity, and code thresholds the probabilities.

Classify documents through deep patent, retail product, biomedical, and source-code hierarchies using parallel beam search over Choice probabilities.

Classify SEC annual reports into 75 industry groups with one Choice each, then read confidence to decide when to fall back to a broader division.