Use cases › IT / internal ops

Gelato: how the agent works and how to build one like it

Summary: Gelato built an engineering ticket-triage agent that raised assignment accuracy from 60% to 90%, per the source.

Source: https://cloud.google.com/transform/101-real-world-generative-ai-use-cases-from-industry-leaders source page fetched

Why this use case works

Triage is a classification problem with a clear right answer (the team that ended up fixing it), so historic tickets double as training and evaluation data. That makes progress measurable from day one. (Our analysis.)

Agent pattern: Internal ops agent

An internal-ops agent living in a chat tool employees already use (Slack, Teams, WhatsApp), with read access to documentation and write access limited to creating tickets or records. Start read-only and widen permissions as accuracy is proven.

How a smaller team could replicate it

Teams can export a few months of resolved tickets, use them to build and test a routing prompt or classifier, run it in suggest-only mode, then switch to auto-assign once accuracy beats the current manual baseline. (Our analysis, not a claim made by the source.)

Want a rough business case first? Try the AI agent ROI calculator, or ask us to scope it.

Risks and guardrails

Risks: misrouted urgent incidents and drift as teams reorganize. Guardrails: never auto-assign severity-one incidents without a human, retrain or update routing rules whenever team ownership changes, and track overrides. (Our analysis.)

What to measure

Assignment accuracy, reassignment count, time-to-first-response, and engineer overrides.

Related IT / internal ops deployments

Part of the AgentStack Atlas catalog of 52 sourced agentic AI deployments. Spotted an error? Email agentstackatlas@mail.grokbot.com.