Chanl

Chanl Overview

The backend platform for AI agents — tools, knowledge, memory, testing, and monitoring

What Is Chanl?

Every AI agent team builds the same six backend services from scratch: tool execution, knowledge retrieval, persistent memory, quality evaluation, automated testing, and conversation analytics. Chanl provides all of them out of the box — so you can focus on the conversation design that actually differentiates your product.

You keep your voice platform (VAPI, Retell, Bland, Twilio). Chanl is the layer underneath that makes your agents intelligent, testable, and observable.

Build → Test → Monitor

Build Test Monitor Insights Tools & Toolsets Knowledge Base Persistent Memory MCP Server Scenarios AI Personas Scorecards Transcripts Analytics Smart Alerts

Build

Give your agents superpowers. Connect tools (CRM, order lookup, ticket creation), upload a knowledge base (docs, FAQs, policies), and enable persistent memory (customer context across conversations) — all exposed via a single MCP endpoint.

Test

Stop embarrassing AI failures. Create scenarios with AI personas — from polite customers to hostile complainers — and score every response automatically with scorecards. Run hundreds of tests in minutes, not days.

Monitor

Every conversation gets a grade. Full transcripts, real-time analytics, and smart alerts when quality drops or an agent goes off-script. You find out from your dashboard, not from Twitter.

What Chanl Replaces

Without ChanlWith Chanl
Build a tool execution runtimeTools + Toolsets + MCP Server
Build a RAG pipelineKnowledge Base — upload docs, get semantic search
Build a persistent memory storeMemory — customer, agent, and session scoped
Build a quality evaluation systemScorecards — auto-grade every conversation
Build an automated testing harnessScenarios + Personas
Build conversation analyticsAnalytics + Transcripts

How It Works

Your agent platform handles the conversation. Chanl handles everything behind it.

Your Agent Platform Chanl Backend Quality Layer MCP Call Data VAPI / Retell / Bland / Twilio MCP Server Knowledge Base Memory Tools Automated Testing Scorecard Evaluation Real-time Analytics

During a call, your agent queries the knowledge base, reads customer memory, and calls tools — all via MCP. After the call, Chanl transcribes it, scores it against your criteria, and feeds the data into analytics. No extra code on your side.

Get Started

Getting Started

Connect your first agent and run a test in 5 minutes

MCP

Set up the MCP connection that powers everything

Integrations

Platform-specific setup guides for VAPI, Retell, Bland, Twilio

Explore by Section

Build

Tools, knowledge base, memory, toolsets — give your agents real capabilities

Test

Scenarios, personas, simulations — catch failures before customers do

Monitor

Transcripts, scorecards, analytics, alerts — see everything your agents say

On this page