Skip to main content
Build your own autonomous AI agent in Base44. Think of it as a personal assistant that monitors activity, runs tasks, connects to tools, and helps manage your workflows automatically. Your AI agent acts like a persistent digital teammate. It can chat with you, use knowledge to reason about problems, run tasks automatically, connect to external tools, and take action across your systems and workflows.
Building your personal AI agent.
AI Agent is currently rolling out and may not yet be available in all Base44 workspaces.

What your AI agent can do

Your AI agent can perform a wide range of tasks, including:
  • Monitor systems and respond to events automatically.
  • Run scheduled tasks and recurring workflows.
  • Connect to external tools such as Slack, Google Calendar, and GitHub.
  • Communicate through messaging channels.
  • Access knowledge and stored files to reason about problems.
  • Trigger actions such as sending alerts, updating data, or generating reports.
  • Search the web, gather information, and open live previews while researching topics.

​
Building your AI agent

Start building your agent by defining what you want it to do and configuring the tools it will use. To build your AI agent:
  1. Log in to Base44.
  2. From the Base44 home screen, click AI Agent.
  3. Click Start building your AI Agent.
  4. Start chatting with your new agent and describe what you want it to do. It will suggest tasks, automations, and tools based on your instructions.
Buildingyouraiagent

​
Configuring your AI agent

After creating your agent, configure how it thinks, what information it can access, and what actions it can perform.

​
Chat

Interact directly with your AI agent, guide its behavior, create workflows, and observe how it performs tasks. Your agent can also perform live tasks such as browsing the web, researching information, and previewing results directly in the workspace.
Chattingaiagent
You can also chat with your AI agent on Slack or Telegram.

​
Brain

The Brain defines how your AI agent thinks, remembers, and uses information. Here you configure the identity, memory, context, and reference materials your agent relies on when reasoning and completing tasks.
Unlike traditional chatbots, an AI agent maintains persistent context using identity, user context, and long-term memory. Knowledge provides context the agent can reason about, while the Files tab stores working files and generated outputs.
Skills Skills determine what your agent is allowed to access and use when performing tasks.
  • Built-in services: These are platform capabilities available to every agent by default, such as the Base44 Backend, which provides database access, functions, file storage, and automations.
  • Connectors: Connectors link your agent to external services using OAuth integrations. Once connected, your agent can read data, monitor activity, or trigger workflows using those tools and services. If the agent cannot perform a task, check whether the required connector is enabled.
Aiagentconnectors 2
Connectors allow your agent to:
  • Read and summarize messages from Slack.
  • Monitor Gmail for important updates.
  • Query APIs to retrieve external data.
  • Access documentation or knowledge sources.
  • Run backend workflows and automations.
Knowledge Your agent’s knowledge defines the information it uses to reason and remember context across sessions. It includes several core files that shape how the agent understands and responds to requests.
Defines who the agent is.
  • Name
  • Avatar
  • Personality
  • Communication style
Defines how the agent behaves and makes decisions. This file establishes:
  • Behavioral principles
  • Communication tone
  • Boundaries
  • Decision-making guidelines
Stores context about the person the agent is assisting, like:
  • Name and preferences
  • Timezone
  • Role and responsibilities
  • Relevant working context
Memory entries store important information the agent should remember across conversations, and are used as the agent’s long-term memory, such as:
  • Decisions
  • Preferences
  • Project context
  • Recurring tasks
Upload documents and resources your agent can reference. These might include:
  • Documentation
  • Spreadsheets
  • Datasets
  • Reference materials
Aiagentknowledge
Knowledge files provide reference material your agent can use when reasoning about tasks. For working files and generated outputs, use the Files tab, which acts as the agent’s workspace storage.
Together, these elements give your agent persistent context, allowing it to remember information, understand your goals, and respond more intelligently over time.

​
Tasks

Use tasks to run automated workflows on schedules or when events occur. Your agent can create tasks from updates, send daily or weekly summaries, sync data between systems, monitor incoming messages, and more.
Agentaitasks

​
Files

The Files section acts as your agent’s workspace file system, where it can store documents, datasets, scripts, and other resources it may need while working. Your agent can retrieve these files when completing tasks, generating outputs, or responding to requests. Files stored here are typically used during workflows or generated by the agent, while knowledge files added in Brain → Knowledge provide reference context for reasoning.
Uploading files for your AI agent
Some system files may also appear here, such as the .agents folder used to store internal configuration and MCP connections.

​
Managing your agent settings

Manage your agent’s security, credentials, and credit usage under the Settings tab. Store API credentials your agent can use, run security scans to detect vulnerabilities, and monitor credit usage and billing.

​
Secrets & Keys

Securely store credentials your agent needs to access external services. Secrets are stored as environment variables and can be used by your agent’s backend functions or integrations.
Aiagentsecrets

​
Security

The Security tab helps you monitor your agent for potential vulnerabilities. You can run a security scan to check your configuration and integrations.
Securityscanaiagent 1
Security scans help ensure your agent operates safely when connected to external services.

​
Credits

The Credits tab shows your agent’s usage and billing activity. From here you can monitor how many credits your agent consumes when running automations, interacting with tools, or responding to messages.
Agentcredits

​
FAQs

Click a question to learn more about AI agents.
Yes. AI agents can connect to external tools using connectors or backend integrations, allowing them to interact with services like Slack, Gmail, or Google Calendar.
  • AI agents work for you across your workspace. They can monitor activity across apps, run automations, connect to tools, and assist with operational tasks.
  • App AI agents live inside an app and interact with that app’s users.
Use an AI agent when you want a personal assistant that helps you manage systems, workflows, and information across your Base44 workspace.
Automations run predefined workflows on schedules or triggers. AI agents operate dynamically. They analyze context, respond to messages, and decide which actions to take using the tools and knowledge you provide. Many workflows combine both. An agent can trigger automations or monitor their results.
No. Apps and AI agents are separate creation types.