Skip to main content

​
General Questions

Rilo is an AI-powered workflow automation platform that lets you create complex automation workflows from natural language. Describe what you want to automate, and Rilo’s AI agents will create the workflow for you.
No! Rilo is designed to work with natural language. However, understanding basic programming concepts can help you create more complex workflows. The generated code is Python, but you don’t need to write it yourself.
Rilo generates Python code for workflow blocks. All code runs in secure sandbox environments. You don’t need to write Python yourself - just describe what you want in natural language.
Yes. All workflow execution happens in isolated E2B sandboxes. Your data is encrypted in transit and at rest. Rilo never stores sensitive credentials - all authentication is handled through OAuth.
Currently, Rilo is only available as a cloud service. Enterprise customers can contact us to discuss custom deployment options.

​
Workflow Capabilities

Rilo can automate a wide variety of tasks:
  • Data scraping and collection
  • Content generation and processing
  • Email and communication workflows
  • Social media posting and monitoring
  • Data analysis and reporting
  • Integration between different services
  • And much more!
Key limitations include:
  • Workflows are DAGs (no circular dependencies)
  • Each workflow has exactly one trigger type
  • Cannot access data from previous workflow runs directly
  • Social scraping only accesses public data
  • Cannot connect to multiple accounts of the same tool in one workflow See Workflow Limitations for details.
No, each workflow execution is isolated. However, you can persist data to external storage (Google Sheets, databases, etc.) and read from those sources in subsequent runs.
No. Each tool integration connects to a single authenticated account. To use multiple accounts, create separate workflows.
Rilo supports all common file types (documents, spreadsheets, images, PDFs, CSVs, etc.). Maximum file size is 10 MB per file.

​
Integrations

Rilo supports 1000+ integrations via Composio, plus custom API integrations and MCP (Model Context Protocol) servers.
Yes! You can add:
  • Custom APIs: Connect to any REST API
  • MCP Servers: Extend Rilo with Model Context Protocol servers Both become available as libraries in your workflows.
MCP (Model Context Protocol) is a protocol for connecting AI applications to external tools and data sources. Rilo supports MCP servers to extend its capabilities.
Most integrations use OAuth. Go to the Integrations page, find the service you want to connect, and click “Connect”. You’ll be redirected to authenticate with the service.
Yes! Workflows can be triggered by webhooks. You can also use webhooks in Wait blocks to pause workflow execution until a webhook is received.

​
Workflow Execution

Workflows execute in secure E2B sandboxes. Each block runs as an independent background task. Data flows between blocks via JSON files.
The workflow execution stops at the failed block. You can see error details in the execution logs and retry the workflow after fixing the issue.
Yes, you can cancel a running workflow. Wait blocks and HITL blocks also pause execution until conditions are met or reviews are completed.
Workflow execution time depends on the complexity and number of blocks. There are no hard time limits, but very long-running workflows may timeout based on sandbox limits.
Blocks that don’t depend on each other can run in parallel. Rilo automatically determines execution order based on block dependencies.

​
Credits and Billing

Credits are consumed when workflows execute. Each block has a base cost of 1 credit, plus additional credits for tool usage (API calls, LLM operations, etc.).
Workflow execution will pause if you run out of credits mid-execution. You can add credits via top-up and resume execution.
No, monthly credits reset on your billing date. However, top-up credits never expire.
Top-up credits are non-refundable but never expire. For subscription refunds, contact support.
AI operations (LLM calls) consume credits based on token usage. Simple operations may cost 1-5 credits, while complex generation tasks can cost 50+ credits.

​
Workflow Limitations

Understanding these limitations will help you design better workflows.

​
1. Single Trigger Type

Each workflow can have exactly one trigger type:
  • Manual trigger
  • Scheduled trigger (e.g., “daily at 9 AM”)
  • App event trigger (e.g., “when new email arrives”)
You cannot combine trigger types (e.g., “daily at 9 AM OR when new email arrives”). Create separate workflows if you need multiple trigger types.

​
2. No Circular Dependencies

Workflows are DAGs (Directed Acyclic Graphs). Blocks cannot depend on each other in a circular way. Each block must have a clear execution order.

​
3. Data Persistence

Workflows cannot access data from previous runs directly. Each execution is isolated. To persist data between runs:
  • Write to external storage (Google Sheets, databases, file storage)
  • Read from those sources in subsequent workflow runs

​
4. Social Scraping - Public Data Only

Social scraping tools can only access public data:
  • ✅ Public posts, tweets, Instagram posts
  • ❌ Private accounts, DMs, restricted content
This is for privacy and terms-of-service compliance.

​
5. Single Account Per Tool

You cannot connect to multiple accounts of the same tool in one workflow:
  • ❌ Fetch from Gmail account A and send from Gmail account B
  • ❌ Read from Google Sheet in workspace A and write to Sheet in workspace B
Solution: Create separate workflows for each account/workspace.

​
6. Security Restrictions

Rilo never creates blocks that:
  • Access environment variables or system configuration
  • Read API keys, secrets, tokens, or credentials
  • Execute shell commands or system calls
  • Access system files
  • Send data to arbitrary external URLs
All API authentication is handled automatically by the platform.

​
Troubleshooting

Check:
  • Do you have enough credits?
  • Is the workflow in “verified” state?
  • Are all required integrations connected?
  • Check the execution logs for error messages
  • Ensure you have the correct permissions
  • Try disconnecting and reconnecting
  • Check if the service requires additional OAuth scopes
  • Contact support if the issue persists
  • Review the block instructions - they may be unclear
  • Check if all required config fields are set
  • Verify that the required tools are available
  • Try regenerating the code with more specific instructions
  • Review which blocks are using the most credits
  • Optimize AI operations (use simpler models when possible)
  • Consider caching results when appropriate
  • Check for unnecessary API calls

​
Getting Help

Need more help? Contact us at [email protected] or check out our documentation.
  • Documentation: Browse our guides and feature documentation
  • Support Email: [email protected]
  • Feature Requests: We’re always improving based on user feedback
Still have questions? Check out our Features documentation or Integrations guide.