Sherlock Calls

TL;DR: Sherlock Calls is an AI SRE/QA analyst that lives in your Slack and reconstructs failed voice calls into case files: correlated timeline, root cause with evidence, and suggested fixes. No dashboards. Full description: Debugging voice AI calls is cross-vendor chaos. When a call fails — drop, timeout, silence, transfer failure — the evidence is split across Twilio call events, ElevenLabs audio and TTS behaviour, webhooks, and CRM records. None of the systems know about each other. Most teams end up doing this manually: pull logs from each platform, try to reconstruct a timeline, guess at the root cause. Sherlock correlates everything into a single incident case file posted in Slack: • Timestamped cross-provider timeline (what happened, where, when) • Webhook correlation — ties call events to the webhooks and tools that fired • Latency attribution — telephony setup, TTS streaming, downstream tool calls • Root cause hypothesis with evidence layering • First checks in triage order with verification steps. Integrations: Twilio, ElevenLabs, Vapi, Retell AI, Genesys, Amazon Connect, HubSpot, Aircall. OAuth only — no agent installation, no code changes. Free tier: 100 credits per Slack workspace. Team plans from $50/month. Best for: Voice AI operators, telephony engineers, on-call SREs, technical founders running voice AI in production