# Examples

> Worked examples for the Decision API: ticket routing, escalation by cost, scam checks, agent tool calls, retrieval relevance, supported claims, prompt injection and contract clauses.

Section: Examples. Web page: https://www.launchables.ai/docs/examples

Short worked examples, each built the same way: the situation, the request, how to read the answer and what to watch once it runs. Copy one, change the state and the options to fit your case, and set thresholds on your own data as described in [Acting on the answer](https://www.launchables.ai/docs/thresholds).

Probabilities and token counts in the example answers are illustrative.

| Example | Question types |
|---|---|
| [Route a support ticket](https://www.launchables.ai/docs/examples/ticket-routing) | choice, noul and score in one request |
| [Hand over to a person or a larger model](https://www.launchables.ai/docs/examples/escalation) | choice, by expected cost |
| [Is this message a scam?](https://www.launchables.ai/docs/examples/scam-check) | noul |
| [Should an agent's tool call run?](https://www.launchables.ai/docs/examples/tool-call) | choice |
| [Does this passage answer the question?](https://www.launchables.ai/docs/examples/retrieval-relevance) | score |
| [Is every claim in a draft supported?](https://www.launchables.ai/docs/examples/supported-claims) | noul, one per claim |
| [Does this text try to take over an agent?](https://www.launchables.ai/docs/examples/prompt-injection) | noul |
| [Does a contract clause meet our policy?](https://www.launchables.ai/docs/examples/policy-clause) | choice |