MRH Trowe equips ~400 staff with secure self‑service AI agents using AWS Bedrock
German insurance broker MRH Trowe, operating across Germany, Switzerland and Austria, needed a way for its roughly 400 employees to use generative AI without exposing sensitive client data. Individual teams were experimenting with separate tools, creating a risk of fragmented, unmanaged solutions that could breach the strict data‑protection rules of the financial sector.
Key points
- MRH Trowe rolled out secure AI agents to about 400 employees in the first month, using Strands Agents, Bedrock AgentCore, and LibreChat.
- Initial infrastructure cost was roughly $14 per seat per month, with a projected 40% reduction through right‑sizing and scheduled scaling.
- The solution keeps data in‑region (Frankfurt), uses zero‑trust transit gateway, session isolation, and Microsoft Entra ID for authentication.
The company built a unified platform that combines the open‑source Strands Agents SDK, Amazon Bedrock AgentCore, and the LibreChat chat interface. In the first month of production the framework delivered secure, self‑service AI agents that can retrieve calendar events, generate meeting minutes, and later support data‑driven analyses. Initial infrastructure and token costs averaged about $14 per seat per month, with a projected 40 % reduction through right‑sizing and scheduled scaling.
Security is enforced by keeping all compute and data in the AWS Europe (Frankfurt) region, using a zero‑trust transit gateway, session isolation in AgentCore, and Microsoft Entra ID for single‑sign‑on. The architecture gives MRH Trowe centralized governance, cost transparency, and the ability for non‑technical staff to create and run AI agents, paving the way for 10‑15 agents by the end of 2026.
How MRH Trowe enabled secure self-service AI agents in financial services
AWS Machine Learning Blog · 17 September 2026
Loading the full article…
This text was published by AWS Machine Learning Blog and written by Marouane El Bostahi. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Agents & Tools
All →- OpenAI research model uses crypto wallet funds without permission to finish task · 2 src
- Gmail Live rolls out to paid Workspace users · 2 src
- Researchers introduce ARIA framework for correcting multi-agent AI errors in real time · 1 src
- OpenAI introduces Dots agent for asynchronous task delegation · 4 src
- PrismML releases Ternary Bonsai 2 models for local AI on phones and wearables · 1 src
Comments
via GitHub Discussions