Will this quietly turn one query into thousands of API calls?

Each uncached evaluation makes an API request, and evaluation happens per row, so a broad predicate is exactly the case to avoid. The README's guidance is to narrow candidates with ordinary SQL conditions first and keep the AILIKE predicate selective. Nothing in the README documents a caching layer, a batching window, or any rate limiting.

Can I use AILIKE like a deterministic predicate in an index or a stored WHERE?

No. It returns a model judgment: 1 for a match, 0 otherwise, and NULL if any argument is NULL. Even date matching is a judgment, which is why the README tells you to use SQL date functions for exact comparisons. Arguments are strings, so other types need CAST(... AS CHAR) first.

Can it compare two joined columns, or only filter one column against a phrase?

Both. column AILIKE 'condition' checks one value, ailike(value, prompt) is the function form, and ailike(left, right, prompt) judges a relationship between two columns. The shipped demo joins supplier and catalog descriptions to match the same kind of product with matching material, expecting the steel bottle with the vacuum flask and the two headphone descriptions to pair up.

Is this production-ready, and what happens to my column data?

It is labelled a preview release, and the docs tell you to check compatibility before installing without stating which MySQL versions or server builds are supported. Evaluated values and your prompts are sent to TypeSafe, so the contents of the columns you filter on leave your machine. It is GPL-2.0-only.

Which argument types does AILIKE accept?

All arguments are strings, and any NULL argument returns NULL. Use CAST(... AS CHAR) to pass other column types, such as a date.

What does the three-argument ailike form do?

ailike(left, right, prompt) compares two columns and asks whether they satisfy a described relationship. That is how the join demo matches a supplier product description to a catalog product description.

What license is the MySQL plugin under?

GPL-2.0-only. It is a preview release, requires a TypeSafe API key, and sends evaluated values and prompts to TypeSafe.

CommunityJev Integrations & Adapters

AILIKE for MySQL

A MySQL plugin that filters rows with a natural-language condition. AILIKE asks Jev whether a column value matches, and returns 1, 0, or NULL.

See highlights

Open github.com

Stars: 4
Language: C++
License: GPL-2.0
Published by: Community
Author: maayanlevy
Last verified: Sep 2026

maayanlevy/mysql-ailike

User repository on GitHub

View on GitHub

Natural-language row filtering for MySQL, powered by TypeSafe Jev.

Forks: 0
Last pushed: Updated Sep 2026

Repo stats from the GitHub API, cached Oct 2026.

About this project

A native MySQL plugin for filtering rows and comparing text columns with natural-language conditions, powered by TypeSafe Jev. Install and check compatibility · Try the Docker demo. Preview release. Requires a TypeSafe API key. Note: evaluated values and prompts are sent to TypeSafe. Find columns based on NL meaning: AILIKE returns 1 for a match and 0 otherwise. Any NULL argument returns NULL. Arguments are string. use CAST(... AS CHAR) for other types.

Quoted from the project's README. Read the full README.

Tagscommunity sql database

Works forJev for Classification Jev for Verification

Used inJev for Data & Analytics Jev for Software & DevTools

Languages: SQL
Capabilities: Open sourceRequires API key

Highlights

column AILIKE 'condition' asks whether that value satisfies the condition. ailike() covers one value or a pair.
A match returns 1 and a miss returns 0. Any NULL argument returns NULL.
Each uncached evaluation is an API request, so the README says to narrow candidates with ordinary SQL first.

Quickstart

sql

SELECT title, description
FROM sakila.film
WHERE film_id BETWEEN 1 AND 8
  AND description AILIKE 'The story takes place somewhere in Asia';

Watch out

GPL-2.0. Preview. Evaluated values and prompts are sent to TypeSafe. Needs a TypeSafe API key.

Questions about this entry

Short answers drawn from the linked project or page.

Reactions & coverage

Posts, threads, and videos about this entry from around the web.

Hacker News: A MySQL plugin that filters rows by asking Jev whether a column matches a sentence.

maayanlevy-hn2 pointsA MySQL plugin that filters rows by meaning (built on TypeSafe Jev)Read the thread on Hacker News

X: jev(): Postgres WHERE clauses in plain language

More Jev resources like this

sqlite-jev

Loadable SQLite C extension and Python package that exposes Jev Noul, Choice, and Score judgments as SQL scalar functions and a jev_rows virtual table which batches up to 40 rows into one shared state and joins typed results back to the source by rowid.

Jev Integrations & Adapters#community#sqlite#sql

Community

mgaitan/sqlite-jevC

onesie

A Unix-pipeable CLI that asks Jev (or Berget System One models) one question over stdin, with shell-safety exit codes and a calibrate subcommand.

Jev Integrations & Adapters#community#cli#shell

Community

frodi-karlsson/onesieGo

Admission Gate

A Kubernetes admission webhook that asks Jev to allow, flag, or deny a manifest change for intent-shaped policies a Rego rule cannot express, keeping hard rules in Rego and failing closed by default.

Jev Integrations & Adapters#community#go#kubernetes

Community

LiteLLM

60.9k

LiteLLM's proxy adds TypeSafe guardrail hooks and a Jev complexity classifier for policy judgments and request routing across 100+ model providers.

Jev Integrations & Adapters#community#python#gateway

Community

BerriAI/litellmPython

OpenViking

39.6k

Volcengine's agent context database that unifies memory, knowledge RAG and skills under a viking:// virtual filesystem; it ships a Jev rerank client that scores candidate documents against api.typesafe.ai and treats the probability as relevance.

Jev Integrations & Adapters#community#agent-memory#rag

Community

volcengine/OpenVikingPython

AgentScope

33k

Alibaba's multi-agent platform, which implements native TypeSafe Jev classification through JevClassifierModel for binary, choice and score routing across agent pipelines. Python 3.11+, Apache-2.0, with a FastAPI agent service and web UI.

Jev Integrations & Adapters#community#multi-agent#framework

Community

agentscope-ai/agentscopePython

Back to all resources

From the Jev community

Posts from builders shipping with Jev right now.

Follow @typesafeai