From portal-hopping to instant answers: HEMA’s journey with MCP and Amazon Bedrock
HEMA built HAL, an internal AI assistant powered by Amazon Bedrock AgentCore, that replaces a fragmented developer‑portal workflow with instant, governed answers. By leveraging the Model Context Protocol (MCP), HAL injects curated knowledge directly into the model’s context, allowing it to answer queries inside the tools teams already use without exposing AWS credentials on the client side. The result is a secure, low‑latency assistant that can be integrated into existing workflows with minimal changes to the user interface.
⚡ Key Takeaways
- HAL runs on Amazon Bedrock AgentCore, using MCP to supply context‑aware, governed knowledge to the model.
- MCP is employed to embed proprietary data into the model context, eliminating the need for client‑side AWS credentials.
- The integration delivers instant answers within existing developer tools, cutting portal‑hopping time and reducing security risk.
- The assistant is deployed via Bedrock AgentCore’s API, with context passed through the `create_agent` call and managed by IAM roles on the backend.
- WhyItMatters: Engineers can now embed AI assistants into their existing tooling without exposing cloud credentials, improving both security posture and developer productivity.
- TechnicalLevel: Intermediate
- TargetAudience: ML Engineers, RAG Practitioners
- PracticalSteps:
- Use Bedrock AgentCore
Engineers can now embed AI assistants into their existing tooling without exposing cloud credentials, improving both security posture and developer productivity.
✅ Practical Steps
- Use Bedrock AgentCore
Want the full story? Read the original article.
Read on AWS ML Blog ↗