Skip to main content

{altText}

Welcome to our docs!

Portia AI is an open source developer framework. We want to allow any developer to deploy agents that are transparent, steerable and authenticated.
Try out our product and give us some feedback on our Discord channel (↗).

Overview of Portia AI​

The core product accessible in our Github repository (↗) is extensible with our complimentary cloud features which are aimed at making production deployments easier and faster.

With our open source capabilities you should be able to:

  • Ingest a user query and leverage our system prompting to generate a structured plan.
  • Run a plan, invoking tools and tracking the plan run state at every step.
  • Build your own tools.
  • Define clarifications to interrupt a plan run and solicit structured human / machine input when necessary, e.g. to handle required authentication or missing input.

Our cloud offering, which can be enabled from the Portia dashboard (↗) works seamlessly with our open source library. It will give you the ability to:

  • Store and retrieve plan run states and review historical plan runs.
  • Invoke a library of cloud-hosted tools with authentication handled for you, and access tool call logs.

{altText}

Why Portia AI​

We are beginning our journey as a developer framework by focusing on this problem set. Shout if you think we should add to this list 🙏

ProblemPortia's solution

Planning: Many use cases require visibility into the LLM’s reasoning, particularly for complex tasks requiring multiple steps and tools. LLMs also struggle picking the right tools as their tool set grows.

Pre-expressed plans: Our open source, multi-shot prompter guides your LLM to produce an explicit Plan in response to a prompt, weaving the relevant tools, inputs, and outputs for every step.

Execution: Tracking an LLM’s progress mid-task is difficult, making it harder to intervene when guidance is needed. This is especially critical for enforcing company policies or correcting hallucinations (hello, missing arguments in tool calls!)

Stateful, steerable agents: Portia will spin up a multi-agent plan to execute and track its state throughout execution via a PlanRun. Using our Clarification abstraction you can define points where you want to take control of plan runs e.g. to resolve missing information or multiple choice decisions. Portia serialises the plan run state, and you can manage its storage / retrieval yourself or use our cloud offering for simplicity.

Authentication: Existing solutions often disrupt the user experience with cumbersome authentication flows or require pre-emptive, full access to every tool—an approach that doesn’t scale for multi-agent assistants.

Extensible, authenticated tool calling: Bring your own tools on our extensible Tool abstraction, or use our growing plug and play authenticated tool library, which will include a number of popular SaaS providers over time (Google, Zendesk, Hubspot, Github etc.). All Portia tools feature just-in-time authentication with token refresh, offering security without compromising on user experience.

Alright let's roll our sleeves up and get you spinning up them agents 🤖! Next up we'll install the SDK locally and validate your setup, before creating a Portia cloud account.

🎉 I'm feeling lucky

If you prefer to just dive right into an example, why not head over to the intro example in our examples repo (↗).

Setup and configuration​