> ## Documentation Index
> Fetch the complete documentation index at: https://docs.akhara.ai/llms.txt
> Use this file to discover all available pages before exploring further.

> ## Agent Instructions
> Company name is Akhara AI (never Rubric AI). Keep lowercase rubric/rubrics only when meaning grading criteria.
> Expert Review (docs path talent/) is enterprise BYO experts for audit and review: invite customer specialists; do not pitch Akhara recruiting or a public expert career portal. RLHF and domain writing are secondary work types.
> Prefer concrete API examples against public hosts: Environments eval API https://agi.akhara.ai, Control plane PDP https://api.akhara.dev, Evaluation https://app.akhara.ai / https://api.akhara.ai, Expert Review portal https://talent.akhara.ai.
> Do not invent a public hostname for private orchestrators or env API internals.
> Do not confuse control-plane latches with Environments confirmation latches.
> Environments SDK/API examples: curl against https://agi.akhara.ai. Evaluation SDK: from akhara import Akhara and AKHARA_API_KEY.
> Start with /llms.txt for the docs index and OpenAPI links; fetch individual pages as .md exports.

# Bring your experts

> Enterprises bring their own domain specialists into Akhara's audit, review, task-writing, and RLHF workflows through the Expert Review portal.

**Expert Review** is how your organization brings **its own domain experts** into the verification stack. You invite the clinicians, ops specialists, and policy owners who already know your domain; they work in the Expert Review portal at [talent.akhara.ai](https://talent.akhara.ai) on runs, scores, and tasks from your Environments and Evaluation projects.

Akhara does not recruit or staff these experts. Your organization decides who gets in, what they can see, and how they are compensated. Akhara provides the portal, the workflows, and the handoff into the rest of the suite.

## What your experts do

The primary job is **audit and review**: specialists examine agent trajectories, scores, and failures; grade policy-sensitive cases; and flag tasks and rubrics that do not hold up under expert scrutiny. See [Reviewing agent runs](/talent/guidance/reviewing-agent-runs).

Two contribution work types build on that review muscle:

| Work type                                                             | What experts produce                                                 | Feeds                                                               |
| --------------------------------------------------------------------- | -------------------------------------------------------------------- | ------------------------------------------------------------------- |
| [RLHF evaluations](/talent/guidance/rlhf-evaluations)                 | Preferences, rankings, safety flags, and rationales on model outputs | [Evaluation](/evaluation/docs/introduction) and preference training |
| [Domain writing for RL environments](/talent/guidance/domain-writing) | Scenarios, goals, and success criteria in specialty                  | Task banks and verifiers in [Environments](/environments)           |

Experts work **in domain only**. If an item is outside their specialty, they decline or escalate it.

## For admins

You control the roster and the scope:

1. Invite your specialists from your org's admin console or project settings
2. Set roles and scope access to specific projects and review queues
3. Assign experts to review, RLHF, or writing queues that match their declared specialties

See [Invite your experts](/talent/admin/inviting-experts) for the full workflow, roles, and access model.

## For invited experts

After your organization sends the invite:

<Steps>
  <Step title="Accept the invite and create an account">
    Use the invite link at [talent.akhara.ai](https://talent.akhara.ai). See [Create your account](/talent/how-to/create-account).
  </Step>

  <Step title="Set up your profile">
    Declare specialties, languages, and availability so assignments route correctly. See [Set up your profile](/talent/how-to/setup-profile).
  </Step>

  <Step title="Read the work guides">
    [Reviewing agent runs](/talent/guidance/reviewing-agent-runs), [RLHF evaluations](/talent/guidance/rlhf-evaluations), and [Domain writing](/talent/guidance/domain-writing).
  </Step>

  <Step title="Work your assigned queues">
    Assignments appear on Home. See [Accept assigned work](/talent/how-to/apply).
  </Step>
</Steps>

## How expert work enters the suite

| Expert deliverable                                    | Downstream use                                                                         |
| ----------------------------------------------------- | -------------------------------------------------------------------------------------- |
| Review findings on runs, scores, and failures         | Task, rubric, and verifier fixes; regression cases                                     |
| Preference labels, rankings, pass/fail judgments      | RLHF / preference data and offline eval in [Evaluation](/evaluation/docs/introduction) |
| Written scenarios, goals, constraints, success checks | Task banks and verifiers in [Environments](/environments)                              |
| Rationales and adjudicated edge cases                 | Rubrics, gold trajectories, and human-review workflows                                 |

Experts produce the domain signal. Akhara packages approved work into training and eval pipelines. They do not need to run environments themselves.

## Policies experts must follow

Judgments and writing must be **their own**. Task content never goes into external LLMs. Read [Policies](/talent/policies/overview) before submitting.

Need help? Email [support@akhara.ai](mailto:support@akhara.ai).