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

# Task handover (early access)

> Early access: move each task to your own AI only after it passes a test, move it back if quality slips, and keep a record your regulator can check.

> [!NOTE]
>
> The open-source framework is live today. The handover is in early access, and we are taking on our first design partners. [Talk to us about early access](https://praison.ai/enterprise/).

Each task moves to your own AI only after it passes a test. If quality slips, it moves back automatically, and every step is recorded for regulators.

```mermaid theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
graph TD
    W[Watch: save each request<br/>and the answer it got] --> S{Split the saved work}
    S -->|70% study pile| O[Your own AI<br/>as it is, or trained]
    S -->|30% exam pile| T[Test: your own AI<br/>takes the exam]
    O --> T
    T -->|not yet| O
    T -->|reaches the pass mark,<br/>e.g. 99%| P[Practise on live work<br/>+ a named person approves]
    P --> H[Hand over<br/>10% trial, then all]
    H -->|quality below 95%,<br/>errors, or a change| B[Roll back<br/>to today's AI]
    B --> W
    W -.-> R[(Record:<br/>sealed diary)]
    T -.-> R
    H -.-> R
    B -.-> R

    classDef agent fill:#8B0000,color:#fff
    classDef tool fill:#189AB4,color:#fff

    class W,T,H,B agent
    class S,O,P,R tool
```

## How it works

1. **Watch:** records how each task is done today.
2. **Test:** your own AI takes an exam made from real past work it never studied. The work is split first: 70% to study, 30% kept back as the exam, never mixed. It must reach a pass mark your risk team sets in advance, such as 99%, proven on at least 400 exam cases.
3. **Hand over:** after 100 practice answers and a named person's approval, a 10% trial with every answer checked, then all of it.
4. **Roll back:** if the checked answers drop below 95%, errors pile up, or anything changes, the task moves back automatically.
5. **Record:** every step goes into a sealed diary with no customer text, which your regulator can check.

The numbers are the handover's strict defaults. No regulator sets one accuracy number for AI: the EU AI Act asks for an "appropriate level of accuracy", and UK bank supervisors ask the board to set its own thresholds. So your risk team sets the pass mark for each task in advance, and it is written in the record.

## Where your own AI comes from

Your own AI is an open model, such as Llama, Qwen or Gemma, running on your own computers. There are two ways to get one:

> [card] [Use an open model as it is](https://praison.ai/docs/features/local-models.md)
>
> Serve it with Ollama, LM Studio, llama.cpp or vLLM. Available today.

> [card] [Train one on your own past work](https://praison.ai/docs/features/praisonai-train-package.md)
>
> PraisonAI Train fine-tunes a small open model on your own NVIDIA GPU servers. Turning watched work into training data is part of early access.

Either way, the model is only a candidate. It takes over a task only after it passes the exam: real past work it never studied, scored against the answers your business already accepted. Your team starts each training run; nothing trains automatically.

> [!WARNING]
>
> Answers written by another AI provider's model can be covered by that provider's terms of use, which may forbid training another model on them. Having your reviewers check those answers does not remove the terms. Check your provider's terms before you train. The safer source is answers your own people wrote or your own open models produced.

> [!NOTE]
>
> **What is "your own AI"?** An open model, such as Llama, Qwen or Gemma, running on your own computers ([fully offline](https://praison.ai/docs/guides/run-fully-offline.md), on [local model servers](https://praison.ai/docs/features/local-models.md) you host). Use it as it is, or train it on your team's past work with [PraisonAI Train](https://praison.ai/docs/train.md) on your own GPU servers (NVIDIA GPU required). Your team decides when to train; nothing trains automatically. Either way, a task moves to it only after it passes its tests.

## Who it is for

Banks, insurers and hospitals that want AI agents but cannot send private data to outside AI services. Keep your data inside, own your AI, and get proof for your regulators.

***

## What it builds on today

The handover is not part of the open-source packages. It is built on parts that are already in the open-source framework, which you can use now:

> [card] [Tracing](https://praison.ai/docs/observability/overview.md)
>
> Record what an agent did, step by step

> [card] [Evaluation with run\_trials](https://praison.ai/docs/features/trials-engine.md)
>
> Run each test case many times and get a pass rate

> [card] [Fine-tuning with praisonai-train](https://praison.ai/docs/features/praisonai-train-package.md)
>
> Train a model of your own on your own data

> [card] [Local models](https://praison.ai/docs/guides/run-fully-offline.md)
>
> Run agents fully offline on your own computers

> [card] [Model router](https://praison.ai/docs/features/model-router.md)
>
> Choose which model handles each task

> [card] [Local model servers](https://praison.ai/docs/features/local-models.md)
>
> Ollama, LM Studio, llama.cpp and vLLM

***

## Early access

We are taking on our first design partners. If you run a highly regulated team, such as a bank, insurer or hospital, and want to try the handover, [get in touch](https://praison.ai/enterprise/).

> [accordion] Is the handover available to install?
>
> No. It is in early access and is not part of `pip install praisonai`. Everything else on this site is open source and available today.

> [accordion] Does PraisonAI train a model automatically?
>
> No. Training is optional, and your team chooses when to run it. You can also use an open model as it is.

> [accordion] Can I build something like this myself?
>
> Yes. Tracing, evaluation, fine-tuning and local models are all in the open-source framework. The cards above link to each one.

> [accordion] Does this change the open-source framework?
>
> No. PraisonAI is still a general-purpose agent framework, and every feature works as before. See the [Introduction](https://praison.ai/docs/introduction.md).


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.