2. DataMiner DevOps Agents
At a Glance
| Format | Demonstration |
| Participants create | Nothing |
| Key message | Intent becomes implementation – the value is accelerated solution creation, not generation |
| Duration | to be defined |
Objective
Show how a clearly stated operational intent turns into a working solution with DataMiner DevOps Agents.
| Intent → | What to show |
|---|---|
| Data Model | Objects and fields describing the operational domain |
| Business Logic | Rules, calculations, and workflows |
| Integration | Connections to existing systems and data |
| Guardrails | Validation, permissions, review, and safe boundaries |
What Is Given
| Item | Details |
|---|---|
| Demo environment | A DataMiner System and the DevOps Agents, prepared and tested |
| Demo script | Pre-tested intents and the expected result of each step |
| Fallback recording | A recording of the full demo in case the live demo fails |
Tip
Use one of the verticals from the hands-on sessions as the demo intent (for example a Media & Broadcast Event model). Participants then recognize the data model when they deploy the backend package in the next session.
Before the Session
- [ ] Run the full demo end to end on the demo environment
- [ ] Reset the demo environment to its starting state
- [ ] Have the fallback recording ready to play
- [ ] Check network and screen sharing
Session Flow
| # | Trainer | Participants |
|---|---|---|
| 1 | State the operational intent as a business outcome, not as a technical request | Listen |
| 2 | Intent → Data Model: show the agent proposing the model; review it out loud | Watch |
| 3 | Intent → Business Logic: show logic being generated and explain the review step | Watch |
| 4 | Intent → Integration: connect the solution to data or systems | Watch |
| 5 | Intent → Guardrails: show validation, permissions, and human review | Watch |
| 6 | Summarize: what would this have taken without agents? Where did time-to-value improve? | Ask questions |
Do's and Don'ts
| Do | Don't |
|---|---|
| Emphasize speed from intent to a reviewed result | Suggest that generated output needs no review |
| Show the human decisions along the way | Live-debug failures – switch to the recording |
| Use pre-tested intents | Improvise untested prompts |
| Link back to value streams and time-to-value | Present it as a code-generation trick |
Success Criteria
- Participants understand that the value is acceleration from intent to a working, governed solution
- Participants see that guardrails and review are part of the process
Transition
"The backend you'll use next was set up the same way. Now it's your turn to turn an operational problem into an app."