Reported operating use · IT service management
AI IT service requests: Redpanda example
Redpanda reports building a Teams assistant for an unnamed customer's IT service desk. It retrieves permitted knowledge, gathers missing form details and submits requests through existing ITSM workflows under the employee's own identity. People keep incident closure and existing approvals.
Prepare IT service requests in Teams while preserving employee identity
Unnamed enterprise customer reported by Redpanda
Reported customer use with first-week observations; customer identity and rollout coverage are not published.
A published third-party example, not a Nuvepro customer case or evidence of a partnership.
The workflow at a glance
Prepare service requests under employee identity
Illustrative summary of the published account
Reported work components. Their arrangement does not imply a fixed execution order or simultaneous work.
- Interaction
Teams assistant gathers request details
The assistant retrieves permitted knowledge and missing form information.
- Integration
Validated fields and record read-back
A custom integration submits through existing ITSM workflows and verifies saved records.
- Authority
Employee identity and existing approvals
Writes retain employee identity; people keep approval and incident-closure decisions.
What the source reports
Redpanda reports building a Teams assistant that retrieves permitted knowledge, gathers missing form details and submits requests through existing ITSM workflows. A custom integration validates fields and reads saved records back. Employee identity is retained for writes; people retain incident closure and existing approvals.
Reported technology: Redpanda Agentic Data Plane. Scope and outcomes are source-reported.
What the enterprise environment would need
The following is a proposed project design for an enterprise doing similar work. Published details are marked; the remaining systems, data and scenarios would be agreed for your project.
Proposed practice environment
Create and verify a fictional service request
Illustrative project design to scope with your enterprise.
- 1 · Set up the environment
Test knowledge and dynamic form definitions
Use a disposable record store and simulated employee identity.
- 2 · Preparation and agent contributions
Assistant prepares the request
Gather required fields and acknowledgments; retain request identity through partial failures.
- 3 · Independent checks
Independent saved-state verification
Compare submitted and saved values; test restricted articles and changed required fields.
- 4 · People decide acceptance
Service owner reviews boundaries
Accept field fidelity, recoverability and employee scope while retaining closure and approval.
A reviewable evidence package
A verified fictional ticket with explicit unverified fields and recovery state.
If a check fails → withhold acceptance.
Creation succeeds but an update fails. Keep the identifier and unresolved fields; do not create a duplicate.
Failure or owner rejection returns to review
- Hold
- → 2 · Revise work
- → 3 · Rerun independent checks
- → 4 · Owner review
Operating boundary. Simulated identity only; no incident closure or approval bypass.
- Systems and evidence
- Proposed: use fictional knowledge articles, dynamic form definitions and a disposable service-management record store.
- People and agents
- Proposed: an assistant prepares requests; employees supply acknowledgments and service owners retain approval and closure decisions.
- State and permissions
- Proposed: scope writes to the simulated employee identity and separate verification reads. Retain request identifiers after partial failure.
A proposed simulation
An illustrative practice design derived from this work, not an executed assessment.
Proposed: prepare a fictional hardware incident and verify its saved state without bypassing approval.
Review gate: A reviewer confirms field fidelity, identity boundaries and recoverable handling of partially completed requests.
Challenge the result: The ticket is created but a follow-up update fails. Preserve its identifier, report unverified fields and avoid creating a duplicate.
What needs checking
- Proposed: compare submitted fields with the saved record using an independent verifier.
- Proposed: test restricted articles, changing required fields and explicit acknowledgment before submission.
- Proposed: refuse incident closure and approval bypass, and expose any record fields that cannot be verified.
Review environment fidelity, human capability and agent performance separately. The enterprise’s accountable owners decide acceptance of the shared workflow and any operating use.
Questions about this example
- What does Redpanda's service desk assistant use AI for?
- Redpanda reports a Teams assistant that retrieves permitted knowledge, gathers missing form details and submits IT service requests through existing ITSM workflows. A custom integration validates fields and reads saved records back to check them.
- Where do people stay involved in this service desk?
- Redpanda says employee identity is retained for writes and that people retain incident closure and existing approvals. It also describes validation of fields and read-back of saved records, though which controls were enabled for this customer is not fully stated.
- How strong is the evidence for this example?
- It is reported customer use with first-week observations, published by Redpanda. The customer is not identified and rollout coverage is not published. The first-week result is not a general service-desk performance guarantee, and no independent evaluation is provided.
- What would an enterprise need to build something similar?
- This is Nuvepro's proposed design, not Redpanda's. It would need fictional knowledge articles, dynamic form definitions and a disposable record store, with writes scoped to a simulated employee. An independent verifier would compare submitted fields with saved records.
Shape this around your enterprise
Bring a similar workflow and the people responsible for it. Together, define the records and systems to use, the result your team will build or configure, and how its owners will check it. Nuvepro scopes the simulation, environment and preparation. Agree any existing-agent evaluation, implementation assistance and production responsibilities separately.
Sources and limitations
Curation review: . This is the date this source example was reviewed, separate from source publication dates.
- This vendor account does not identify the customer or provide an independent evaluation.
- The first-week result is a provider-reported observation, not a general service-desk performance guarantee.
- The article separates available platform controls from controls actually enabled for this customer; do not assume every described policy was deployed.
- Proposed practice and checks are illustrative designs, not an executed Nuvepro demonstration or a replica of customer systems.
- How we built an AI assistant for an enterprise IT service deskSource publication: 2026-10-27