docs: add Tuning Engines AI gateway example

[cerebrixos]

Adds a small TypeScript integration example showing a Restate service handler calling Tuning Engines as an OpenAI-compatible governed AI endpoint.\n\nThe example uses placeholder environment variables only and keeps Restate responsible for durable execution while the gateway handles model routing, policy checks, and trace correlation.\n\nValidation run locally:\n- npm install --ignore-scripts\n- npm run typecheck

[github-actions]

All contributors have signed the CLA ✍️ ✅
_{Posted by the CLA Assistant Lite bot.}

[cerebrixos]

I have read the CLA Document and I hereby sign the CLA

[cerebrixos]

recheck

[cerebrixos]

Hello team @joebowbeer . Please do consider this PR for merge. So much appreciate in advance :)

[gvdongen]

Hi, thanks for contributing! I didn't hear about Tuning Engines yet. Could you provide a bit of extra information like a website or so?
Regarding this PR. This repo is for general-purpose examples (not AI ones). Our AI examples live here: https://github.com/restatedev/ai-examples
We don't yet have the structure there to support listing LLM SDKs. That's something that is on the roadmap.
Also, the code doesn't use the Restate SDK for the non-deterministic steps. Both the ID generation and the LLM call need to be wrapped into durable steps, otherwise replay doesn't work.
If you want, you can open a PR in the other repo, and once we have the structure in place, we will merge it.
Thanks again!