JAMS JAX

JAX is the AI agent built into the JAMS Web Client, the enterprise job scheduler from JAMS Software. Operations and IT teams use it to find jobs, troubleshoot failures and get documentation-grounded answers in plain language, with every write action held for user approval and the AI model chosen by the customer.

JAX is an AI agent that runs inside the JAMS Web Client, the browser interface for the JAMS workload automation and job scheduling platform. JAMS Software shipped it in JAMS V7.9.0 (release notes dated July 9, 2026) and announced general availability on July 24, 2026, alongside JAMS MCP, a Model Context Protocol server that exposes the same scheduler to Cursor, VS Code with Copilot, Claude Code, Claude Desktop and Codex. According to the release notes, JAX lets users ask questions, troubleshoot errors, submit jobs and configure JAMS in natural language. Answers are grounded in the JAMS user guide and checked against a built-in glossary. JAX does not host a model: the customer picks one, either a commercial provider such as OpenAI or Anthropic or a local model running entirely inside its own network, and JAMS states it never trains on customer data. The agent acts as the signed-in user with that user's exact JAMS permissions, reads proceed freely, write actions pause for explicit approval, and all operations are written to dedicated logs and the JAMS audit trail. JAX acts only when asked, does not learn between sessions and does not retain conversations on the server. Both JAX and JAMS MCP are included in JAMS Web with no separate licence, SKU or extra cost. JAMS itself dates to 1987, was owned by Fortra until PSG, 2ndWave Software and employees agreed to buy it in April 2025 (together with Skybot), and the company says it serves more than 850 customers. Peer reviewers place it alongside Control-M, AutoSys and Redwood's ActiveBatch and RunMyJobs.

Category

Automation & Workflows

Pricing

Subscription, Contact for pricing

Target Market

IT Operations, CIOs, DevOps Engineers, SREs, System Administrators

Deployment

Self-hosted, Hybrid

Founded

1987

Headquarters

Middletown, Delaware, United States

Customers

850+ (JAMS Software, July 2026, for the JAMS platform)

• ✓
  Natural-language job search

  Finds jobs and runs across the JAMS environment from a plain-language question, so operators do not have to build filters or remember folder paths.

• ✓
  Failure troubleshooting

  Diagnoses failed jobs in plain language, which shortens the time from an alert to understanding what broke and where.

• ✓
  Documentation-grounded answers

  How-to answers are grounded in the JAMS user guide and checked against a built-in glossary, reducing generic or invented guidance from the model.

• ✓
  Bring your own model

  Customers choose a commercial provider such as OpenAI or Anthropic or a local model, so the AI can run with no external service involved.

• ✓
  Approval gate and audit logging

  Reads run freely, write actions pause for explicit user approval, and every operation is recorded in dedicated logs and the JAMS audit trail.

• ✓
  User-scoped permissions

  JAX acts as the signed-in user with that user's exact JAMS permissions, with no elevated AI service account to secure separately.

• ✓
  JAMS MCP companion

  A Model Context Protocol server exposes the same job queries and run controls (submit, hold, release, reschedule, cancel) to Cursor, VS Code with Copilot, Claude Code, Claude Desktop and Codex.

• •
  Overnight batch failure triage

  An on-call operator asks JAX why a job failed overnight and gets a plain-language diagnosis inside the Web Client instead of reading raw logs.

• •
  Finding jobs in a large estate

  An administrator describes a job by what it does and JAX locates it across folders and agents, saving time spent building queries by hand.

• •
  Approved run control

  A user asks JAX to submit or reschedule a job, reviews the proposed action, and approves it, with the change recorded in the JAMS audit trail.

• •
  Air-gapped or data-sensitive AI use

  A regulated organisation points JAX at a local model inside its own network so job names, logs and schedules never reach an outside AI provider.

• •
  Scheduler access from the IDE

  An integration developer uses JAMS MCP from Claude Code or Cursor to check run status and investigate a failure without switching to the scheduler console.

• •
  Onboarding new operators

  New staff ask JAX how-to questions about JAMS features and get answers grounded in the official user guide, reducing reliance on senior admins.

JAX has no separate price: it ships with JAMS Web at no extra licence, SKU or cost. JAMS itself is licensed per environment with unlimited executions; the vendor lists Core from $833/month billed annually (Capterra lists Core at $9,996/year and Advanced at $19,996/year), and Advanced is quote-based. Model costs depend on the provider you connect, and a local model adds no external AI fee. Some PeerSpot reviewers complained about price increases under Fortra ownership.

This page was written from 8 sources, 4 on domains other than jamsscheduler.com.

1. 1.jamsscheduler.com — aivendor
2. 2.jamsscheduler.com — mcp servervendor
3. 3.jamsscheduler.com — release notesvendor
4. 4.jamsscheduler.com — pricingvendor
5. 5.prnewswire.co.uk — jams launches ai for enterprise job scheduling jax and jams
6. 6.capterra.com — JAMS
7. 7.peerspot.com — jams reviews
8. 8.finsmes.com — psg and 2ndwave software acquires jams from fortra