Launch HN: Voker (YC S24) – Analytics for AI Agents
90% of AI agent teams only know about failures when customers complain. Voker just shipped the fix.
Why it matters
Agent analytics is an emerging category within the AI product stack. Voker addresses a real gap: traditional observability tools and eval frameworks don't give product teams visibility into real-world agent performance trends and user intent patterns at scale.
The key facts
8 to knowYC S24 cohort company
90%+ of surveyed YC founders rely on customer complaints to detect agent failures
Core primitives: Intents, Corrections, Resolutions for conversational agent analysis
LLM stack agnostic SDK (OpenAI, Anthropic, Gemini support)
Free tier: 2,000 events/month
Paid plans start at $80/month with 30-day free trial
Uses LLMs for annotation but data engineering (stats, classification) built on deterministic systems to avoid LLM hallucination in analytics
Python and TypeScript SDKs available
Go to the source
Hacker Newsvoker.ai
Publisher excerpt: Hey HN, we're Alex and Tyler, co-founders of Voker.ai (https://voker.ai/), an agent analytics platform for AI product teams. Voker gives full visibility into what users are asking of your agents, and whether your agents are delivering, without having to dig through logs. Our main product is a…