Mahal Systems EnterpriseOps-Gym (NOWAI-Bench)
Fits ITSM, HR and customer-service agents that write to a system of record. Borrow the SQL verifiers on final database state (about 5 conditions per task, all must pass) and the one-container-per-domain MCP setup.
What it asks the agent to do
- resolve an ITSM incident or request by updating records
- process an HR case (leave, onboarding, records)
- handle a customer service case
- act on email, calendar, drive and team chat data
- chain actions across several of these domains
What to borrow from it
grading designsandbox designtaxonomybaseline
Caveats
Fully synthetic data; 40% of tasks are held out; built by a vendor whose platform the tasks resemble.
Details
| Workloads | it ops, hr, customer support, email calendar, workflow automation |
|---|---|
| Interaction | single turn |
| Environment | containers |
| End state checked | Yes |
| Tools | 512 MCP tools over 164 database tables |
| Grading | state diff, programmatic check |
| Size | 1,150 |
| Data | synthetic |
| Personal data | none |
| License | Apache-2.0 |
| Harness | own repo |
| Status | active |