# Voker

> Voker is an analytics platform for AI agents that tracks intents, corrections, and resolutions to help product teams measure and improve agent performance.

Voker is an agent analytics platform built by Invoke Labs, Inc. that transforms AI agent conversation data into structured, actionable insights for product and business teams. It integrates via a lightweight Python or TypeScript SDK and begins populating a dashboard with performance data after just a few lines of code. The platform is designed for teams running high-volume conversational AI products who need visibility beyond raw traces.

## What It Is

Voker sits in the observability and analytics layer of the AI agent stack. Rather than surfacing low-level LLM traces, it automatically classifies three core signals from every conversation: **intents** (what the user wants), **corrections** (moments where the user signals the agent got something wrong), and **resolutions** (whether the agent ultimately solved the user's goal). These signals are aggregated into dashboards that product managers, analysts, and business stakeholders can use without engineering support.

## How the Core Analytics Work

The platform processes multi-turn conversations and applies automatic classification to extract structured metrics:

- **Intent detection** — identifies user goals from natural language across every session
- **Correction tracking** — surfaces friction points before they become churn, flagging when users repeat or rephrase requests
- **Resolution measurement** — tracks success rates per intent category across all interactions
- **User behavior insights** — alerts teams when users become frustrated or abandon sessions ("rage-quit")
- **Performance tracking** — quantifies the impact of agent changes over time and supports rollback decisions

The homepage dashboard example shows metrics including total sessions, correction rate, resolution rate, and intent category breakdowns with trend lines.

## Integration and Stack Compatibility

Voker is designed to work alongside existing observability tools rather than replace them. The homepage explicitly lists compatibility with Langfuse, LangSmith, PostHog, Mixpanel, and Amplitude. Supported LLM providers and frameworks include OpenAI, Anthropic, Gemini, LangChain, CrewAI, and the Vercel AI SDK. Installation requires two lines of code (`pip install voker`) with no infrastructure changes. The platform also supports self-hosted deployment for enterprise use cases.

## Target Audience

Voker targets cross-functional teams building agent-powered products, specifically those with:
- High interaction volume (the site references 1,000+ chat sessions per month as a threshold)
- Complex multi-turn conversations involving tools, RAG, or MCP
- Product managers, analysts, or business stakeholders who need self-service access to agent insights without filing engineering tickets

The platform positions itself as a bridge between engineering-focused trace tools and business-facing analytics, enabling non-technical stakeholders to measure agent ROI and connect conversational metrics to outcomes like conversion and retention.

## Vendor-Published Customer References

The Voker homepage includes testimonials attributed to named executives. Ben Yahalom (CEO of True Classic) is quoted saying the platform made it possible to monitor and optimize AI features. Carlos Moreno (CTO of Dutch) is quoted on optimizing agents in production. Peter Greig (VP of Data & Analytics at Lull) references intent and resolution tracking for product investment decisions. These are vendor-published claims on the product homepage.

## Features
- Intent detection and classification
- Correction rate tracking
- Resolution rate measurement
- Queryable conversation timelines
- Agent performance tracking
- User behavior insights
- Rage-quit and frustration alerts
- Agent auto-optimization (Beta)
- Self-service analytics dashboard
- Lightweight Python and TypeScript SDK
- Self-hosted deployment option
- SSO support (Enterprise)
- Unlimited seats on all plans
- Ecosystem compatibility with Langfuse, LangSmith, PostHog, Mixpanel, Amplitude

## Integrations
OpenAI, Anthropic, Gemini, LangChain, CrewAI, Vercel AI SDK, Langfuse, LangSmith, PostHog, Mixpanel, Amplitude

## Platforms
WEB, API, DEVELOPER_SDK

## Pricing
Freemium — Free tier available with paid upgrades

## Links
- Website: https://voker.ai
- Documentation: https://app.voker.ai/docs/overview
- EveryDev.ai: https://www.everydev.ai/tools/voker