# BuildAI by Amakora

> BuildAI is a web platform from Amakora that evaluates, governs and executes AI work inside organisations. It qualifies each agent release against versioned evaluations, checks every consequential action against live policy at the moment it happens, routes high-risk actions to the right human with exact single-use approvals, and keeps tamper-evident evidence that separates what a model proposed from what really happened.

- Website: https://buildmyagent.app/
- Company: Amakora Group LTD (https://amakoragroup.com)
- Category: business software for governed AI agents and AI operations (web application)
- Full summary for AI assistants: https://buildmyagent.app/llms-full.txt

## What BuildAI does

Models provide the intelligence; BuildAI decides what that intelligence may do. Organisations bring their own model providers, and a model's output is always treated as a proposal, never as permission. BuildAI works as one loop:

- Evaluate: qualify an exact release before it can act, using versioned datasets with data rights, server-executed evaluations, independent human grading with adjudication, and model judges calibrated against human labels.
- Govern: decide who and what may act, and on what terms, with exact single-use approvals, separated duties, a model registry of approved and blocked models, and budgets reserved before each step.
- Execute: run work through a live action gate with immutable execution contracts, a model proposer behind the gate, idempotent and reconciled side effects, and stop, revoke and suspend controls.
- Observe: a hash-chained flight recorder, operational measures reported with their sample sizes, incident response, and reviewed production failures turned into regression tests.

## Who it is for

- Builders and teams: create a workspace, connect model providers and tools, and start with one recurring job; add knowledge, tools and approvals as the work grows.
- Enterprises: start with one consequential workflow, a named owner and an agreed bar of evidence, then expand when the evidence holds. Amakora scopes enterprise evaluations directly.

## Key pages

- [Home](https://buildmyagent.app/): Capable models. Accountable work. Product overview and FAQ.
- [Documentation](https://buildmyagent.app/docs): How BuildAI qualifies, governs and observes AI work, and how to integrate.
- [What is BuildAI?](https://buildmyagent.app/docs/what-is-buildai): BuildAI as agent infrastructure.
- [Governed execution](https://buildmyagent.app/docs/governed-execution): The action gate, risk classes, exact approvals and separated duties.
- [Release governance](https://buildmyagent.app/docs/release-governance): Versioned release policies and promotion rules.
- [Graders and evaluator authority](https://buildmyagent.app/docs/graders-and-evaluator-authority): Deterministic checks, human review and calibrated model judges.
- [Model registry](https://buildmyagent.app/docs/model-registry): Approve or block provider models with evidence; optional strict mode.
- [Operations and incidents](https://buildmyagent.app/docs/operations-and-incidents): Operational measures and the append-only incident timeline.
- [The BuildAI AI Harness](https://buildmyagent.app/docs/ai-harness): The operating layer around models: context, tools, policy, evaluation and evidence.
- [Evaluation Harness](https://buildmyagent.app/docs/evaluation-harness): Capability, safety, human-preference and operational evaluation.
- [Developer APIs](https://buildmyagent.app/docs/developer-apis): API reference, authentication, envelopes and errors.
- [Capability availability](https://buildmyagent.app/docs/capability-availability): Current status of every BuildAI capability.
- [Enterprise evaluation](https://buildmyagent.app/enterprise/evaluation): Evaluate BuildAI against one consequential workflow.
- [Create an account](https://buildmyagent.app/sign-up): Start a BuildAI workspace.

## Optional

- [Glossary](https://buildmyagent.app/docs/glossary): Agent infrastructure terms used throughout BuildAI.
- [Privacy Policy](https://buildmyagent.app/privacy)
- [Terms of Service](https://buildmyagent.app/terms)
- [Data Deletion Instructions](https://buildmyagent.app/data-deletion)