Betting support agents that escalate responsible-gaming worries
Fanatics Betting and Gaming runs a supervisor agent with tools for FAQs, account data and transactions, plus a small classifier that sends high-severity responsible-gaming messages straight to a human. In two months it reports containment up about 56% and resolution up about 53%.
- Agent or tool
- Multi-agent support (Amazon Bedrock)
- Topic
- Customer Support
- Reported on
- Blog (AWS)
- Added
- 2026-10-29
Why it matters
Support is repetitive and time-sensitive. Agents can draft answers, look things up and handle simple requests, so people can focus on the tricky ones. It is one of 21 sightings filed under Customer Support on Agent Sightings (number 10 by interest), and one of 1 involving Multi-agent support (Amazon Bedrock).
How you could try it
Here is a reusable prompt, written by us in our own words, that captures the task. Paste it into Multi-agent support (Amazon Bedrock) or any similar agent and adjust the details to your situation.
Answer customer questions using our state-by-state rules, account and transaction tools. Check every message for signs of problem gambling and hand those to a human straight away with the conversation.
Tip: Start small: run it once on a low-stakes example, read what the agent did, and only then let it run on real work.
Source
Read the original first-hand account on Blog (AWS): https://aws.amazon.com/blogs/machine-learning/how-fanatics-betting-and-gaming-b…. Our summary is a paraphrase; details are in the source.
Featured in these guides
Built an AI agent? Feature your agent on Agent Sightings, free submission or a featured spot.