Open source integration layer for AI agents with secure, multi-tenant support
Filter, extract and score customer enquiries from email and web forms
Stage
Not given
MVP
Founded
Not given
Not given
Based in
Not given
AU
Pricing model
Freemium
Free
Pricing
Hobby: $0/mo with up to 50 connections, 100k webhook events, unlimited managed permissions/auth pages, up to 3 team members, and Discord community support. Pro: $200/mo with unlimited connections, unlimited webhook events, unlimited managed permissions/auth pages, unlimited team members, and Slack s
Not given
Key features
OAuth and multi-tenant credential management
Permissions system with approval workflows
Webhooks and polling for data synchronization
Multiple integration connectors
Self-hosted and cloud-hosted options
Managed authentication pages
Agent-safe credential handling
API caching per tenant
Custom integration scaffolding
Community-contributed integrations
Fast language model classifier with confidence scoring
Custom business rules for filtering
Webhook support with signature verification
Automatic pagination and retry logic
Typed errors with detailed error information
Contact extraction and duplicate detection
Web console for inbox and lead management
Workspace support for agency use
MCP server for AI assistant integration
What makes it different
Fully open source to avoid vendor lock-in
Credentials never stored in Corsair's database
Free self-hosted option with no per-seat pricing
Handles token refresh and API deprecations automatically
Works with or without AI agents
No dependencies in TypeScript SDK — works on any runtime with fetch
Never deletes messages — junk can be rescued with one click
Confidence-based filtering errs on the side of including potential leads
SDK generated from same OpenAPI spec as API reference