JAX AI
JAX AI: an agent built into JAMS
Ask a question. Get an answer grounded in real JAMS documentation, not a guess. JAX AI runs on the model you choose, and it asks before it changes anything.
Watch JAX AI diagnose a real job failure in under two minutes.
The 2 a.m. problem
The answer is always in another system
A job fails, and root cause means digging across consoles while the clock runs, especially on call. New hires stall on the one person who knows the environment cold. JAX AI puts a grounded answer where you already are, so a trivial question stays trivial.
Ask JAMS anything, right where you work
Open JAX AI from any page in the JAMS Web Client. Ask it to find a job, explain a failure, or walk you through a task, in plain language. Every how-to answer is grounded in the JAMS user guide and checked against a built-in glossary.
What JAX AI can do at launch
Read
Find and inspect jobs, see what is running, diagnose a failure, and look up how-to guidance, instantly, in plain language.
Write
Submit a job, create a folder, and manage a run: hold, release, reschedule, or cancel. In the current release, it never edits or deletes the job behind a cancelled run.
Coming in R3 — Roadmap
AI-assisted creation of new jobs and workflows from a plain-language description, gated by the same approvals and permissions as everything else.
Built to grow, not to overreach
An agent with guardrails, not an autonomous one
JAX AI acts only when you ask. It runs as the signed-in user, with that user’s exact permissions, and it pauses before every write for your confirmation. Reads flow freely. Every JAX AI operation is recorded in its own dedicated log, and any change made through the JAMS API lands in the JAMS audit trail like any other change. The action set will grow; this control model will not. JAX AI does not learn between sessions, and your conversations are not retained on the server.
Choose the model JAX AI runs on
Run JAX AI on a commercial provider such as OpenAI or Anthropic, or on a local model entirely inside your own network. JAMS never trains on your data, and with a local model, nothing leaves at all.
Works with any LLM
Built for the person who answers the page
IT ops and batch teams
A fast, trustworthy answer without leaving JAMS.
Support and on-call staff
A grounded answer at 2 a.m., not a search through old tickets.
New team members
Ask JAX AI instead of bottlenecking the one expert who knows the environment cold.
Common questions, direct answers
Is JAX AI an autonomous agent?
Is JAX AI an autonomous agent?
No. JAX AI is an agent with guardrails, not an autonomous one. It acts only when asked, as the signed-in user, and it waits for your approval before any change.
Can JAX AI delete or edit our jobs?
Can JAX AI delete or edit our jobs?
No. In the current release, it cannot delete anything or edit an existing job, folder, schedule, or agent. Cancelling stops a run; the job behind it is never changed.
Does JAX AI learn from our data?
Does JAX AI learn from our data?
No. It does not learn between sessions, and conversations are not retained on the server.
What models does JAX AI support?
What models does JAX AI support?
JAX AI works with OpenAI-compatible providers and local backends such as Ollama or LM Studio. Reasoning or thinking-mode models are not supported directly.
What does JAX AI cost?
What does JAX AI cost?
Nothing extra. JAX AI is part of JAMS Web, with no separate license, SKU, database, or login.
Is everything JAX AI does logged?
Is everything JAX AI does logged?
Every JAX AI operation is recorded in its own dedicated log, and any change made through the JAMS API lands in the JAMS audit trail like any other change.
