Archy
Book a walkthrough

ArchyAgents · Workbench and user interface

ArchyAgents: agent workflows your team can see and control

ArchyAgents is the workbench and user interface for the Archy AI Platform. Use it to design workflows, choose which agents take part and which tools and knowledge they can use, then edit, start and stop each workflow and see what every run did. Workflows can also be started through the platform API or MCP server.

For engineering, platform and operations leads taking agents from experiment to everyday work.

Request accessExplore the platform
Illustration of the ArchyAgents workflow canvas: Intake to backlog runs four steps, two done, one running, one waiting. Panels show an MCP backlog tool, built-in RAG and ArchyKnowledgebase.

How it works

Four stages from idea to an operated workflow.

  1. Define the workflow.

    Describe the goal, the handoffs and the decisions before anything runs.

  2. Connect tools, systems and knowledge.

    Give each agent the tools, MCP integrations, memory and knowledge it needs, and keep that boundary visible to the people responsible.

  3. Start, stop and adjust.

    Start a workflow in ArchyAgents or through the platform API or MCP server, stop it when you need to, and edit it at any time.

  4. Run, inspect and improve.

    Follow what happened in each run and adjust the workflow where it fell short.

Capabilities

What the workbench is designed to cover.

ArchyAgents is the user interface to the agent and workflow services of the Archy AI Platform. Everything below is part of the workbench.

  • Workflow editor

    Lay out the steps, agents and handoffs of a workflow before it runs.

  • Workflow control

    Edit, start and stop workflows in ArchyAgents, and start them through the platform API or MCP server.

  • Agents

    Give each agent one job in the workflow and pass its result to the next step.

  • Tools via MCP

    Let agents use software tools through the Model Context Protocol.

  • Built-in RAG and knowledge management

    Store information and let agents retrieve it with simple retrieval-augmented generation (RAG).

  • ArchyKnowledgebase

    Give agents knowledge from many systems, structured so they understand concepts and relationships, and connect your own RAG systems or vector databases, through advanced RAG pipelines.

  • Agent memory

    Short-term memory for the task in hand, and long-term memory that carries over.

  • Voice interface

    Work with agents by voice.

  • Observability

    Trace each run, keep audit records, and look back over the history of workflow runs.

See ArchyAgents at work

In the ArchyAgents workbench: a workflow in the visual editor, the agents that run it and usage across them.

ArchyAgents workflow editor, Visual view: several agent nodes joined by connecting lines, a branch node and an end node, with an editing and zoom toolbar and a minimap.
Design workflows visuallyLay out a workflow as a graph of agents and the handoffs between them, then edit it on the canvas before it runs.

Part of the Archy AI Platform

The workbench and user interface for the platform.

The Archy AI Platform is about designing and operating AI workflows. ArchyAgents is where you do that: the workbench and user interface for the platform's agent and workflow services. Agents use the platform's built-in RAG and knowledge management, and can draw on ArchyKnowledgebase, the Data and Knowledge Platform, when knowledge comes from many sources or from RAG systems you already run. ArchyAgents is also where you manage the dedicated agents you create for ArchyAssistant workflows in Jira.

See the platformArchyAssistant

How ArchyAgents connects

  • Built-in RAG and knowledge management for storing and retrieving information
  • ArchyKnowledgebase for knowledge from many systems, mapped into concepts and relationships, including your own RAG systems
  • The Archy AI Platform MCP server and OpenAPI, through which other agents and systems reach the platform and start workflows
  • ArchyAssistant: dedicated agents for its Jira workflows, managed here

Enterprise

ArchyAgents in your own environment

For organisations that need it, ArchyAgents can be deployed as PaaS into your environment. Your data stays isolated from other customers, and the product can be customised to your processes and integrations. We scope each deployment with you.

  • PaaS deployment into your environment
  • Full data isolation
  • Customisation for your processes and integrations
Get more informationHow enterprise deployment works

Frequently asked questions

Request access

Have a workflow in mind? Start there.

Describe the process, the tools involved and where people need to stay in control. We will tell you whether ArchyAgents fits today.

Request accessBook a platform walkthrough

More from the Archy AI Platform