Skip to content

Integration

Amazon Bedrock logo

Amazon Bedrock

Harness and evaluate foundation models served through Amazon Bedrock, with Hyperpriors traces and evaluation gates alongside your existing AWS governance.

Request access to connect

Connections are provisioned in private beta — not authorised from this page.

01 — Permissions

When connected, Hyperpriors is limited to the following — declared up front, revocable at any time. Scope is granted during provisioning, not by a button on this page.

  • 01

    Route agent requests to foundation models on Amazon Bedrock through the Hyperpriors harness

  • 02

    Run evaluation suites against specific Bedrock models and model versions

  • 03

    Enforce guardrail policies on Bedrock inputs and outputs at runtime

  • 04

    Record traces, token usage, cost, and latency for every Bedrock invocation

02 — Details

Built by
Amazon Web Services
Website
aws.amazon.com/bedrock (opens in new tab)
Category
Cloud AI platform
Docs
docs.aws.amazon.com/bedrock (opens in new tab)

03 — Notes

About Amazon Bedrock

Amazon Bedrock is a fully managed AWS service that provides access to foundation models from providers including Anthropic, Meta, Mistral AI, Cohere, and Amazon itself, through a single API. It handles inference capacity, offers provisioned throughput for predictable workloads, and supports model customisation. Because it runs inside the AWS boundary, it inherits the account’s existing controls: IAM for access, CloudWatch and CloudTrail for monitoring and audit, and PrivateLink for network isolation. For many organisations, Bedrock is the sanctioned route to production model access.

What the integration does

Hyperpriors treats Bedrock as a model platform behind the harness. Agent requests are routed to Bedrock-hosted models with the retries, timeouts, and fallback behaviour you configure centrally, so switching between models on the platform is a configuration change rather than a code change.

Evaluation suites target specific Bedrock models and versions, letting you compare candidates side by side before promoting one to production. Guardrails inspect inputs and outputs at runtime, and every invocation is recorded with full trace context — agent, step, tools, cost, and latency.

None of this replaces your AWS governance. IAM policies, CloudTrail audit logs, and network controls continue to apply exactly as before; Hyperpriors adds the model-level layer — evaluation gates, behavioural guardrails, and per-call traces — that cloud tooling does not provide.

Get started

Connect your AWS credentials from the Hyperpriors dashboard, or contact us to talk through your deployment.