# Vernix > An AI assistant that joins your video calls, connects to your tools, and answers questions with real business data during meetings. ## What Vernix Does Vernix is an AI meeting assistant that joins Zoom, Google Meet, Microsoft Teams, and Webex calls. It connects to tools like Slack, Linear, GitHub, and CRM systems, then answers questions and takes action during the call using live data. It also transcribes conversations, generates summaries, extracts action items, and provides searchable meeting history. ## Key Features - Tool Integrations: Connect Slack, Linear, GitHub, or your CRM. Ask Vernix during a call and get answers from your connected tools. - Voice Agent: A live voice agent that listens and responds during calls. Say "Vernix" followed by your question. - Silent Mode: Text-only agent that responds via meeting chat. No audio, no disruption. - Meeting Transcription: Real-time, speaker-identified transcription. Searchable immediately. - AI Summaries: Automatic summaries with key decisions after every call. - Action Items: Tasks extracted from conversations and tracked per meeting. - Cross-Meeting Search: Semantic search across all your meetings and documents. - Knowledge Base: Upload PDFs, DOCX, TXT, or Markdown. The agent uses them during calls. ## Pricing ### Free Plan - 5 silent meetings per month, 30 minutes total - 20 AI queries per day - 5 documents, 50MB storage - No credit card required ### Pro Plan - €29/month (or €24/month billed annually) - 14-day free trial with 90 minutes of call time - Voice agent, silent mode, and tool integrations - 200 documents, 500MB storage - 200 AI queries per day - €30 monthly usage credit included - Voice calls: €3/hr, silent calls: €1.50/hr - Credit covers ~10h voice or ~20h silent per month - API access (1000 requests/day) and MCP server/client connections ## Available Integrations ### Communication - Slack: Search messages, channels, and users. Send follow-ups after meetings. - Telnyx: Voice calls, SMS, and messaging APIs. ### Project Management - Linear: Check sprint status, look up issues, and create tasks from meetings. - Atlassian: Access Jira issues, Confluence pages, and Compass services from one connection. - Asana: Track projects, tasks, and team workloads. - Monday.com: Track workflows, items, and team updates. ### Dev Tools - GitHub: Check PRs, review commits, and look up repository activity. - Firecrawl: Scrape and extract content from any website. - Browserbase: Cloud browser automation for AI agents. - Neon: Manage serverless Postgres databases and run SQL queries. - Cloudflare: Manage Workers, KV, R2, DNS, and other Cloudflare services. - Supabase: Manage Postgres databases, auth, and storage. - Apify: Access 4,000+ web scraping and automation actors. - Context7: Up-to-date code documentation for any library. - Figma: Look up design files, components, and comments. - GitLab: Check merge requests, pipelines, and issues. - Sentry: Look up errors, performance issues, and release health. ### CRM - Pipedrive: Access deals, contacts, and pipeline stages during sales calls. - Intercom: Access conversations, contacts, and help center articles. ### Productivity - Notion: Search pages, databases, and wiki content for context during calls. - Airtable: Query bases, tables, and records. - Exa: AI-powered web search with real-time results. - Tavily: AI search engine optimized for research and RAG. - Zapier: Connect 7,000+ apps and automate workflows. ## Coming Soon - HubSpot: Look up contacts, deals, and company data during client calls. - Dropbox: Search files and shared folders. - Discord: Search messages, channels, and server activity. - Zendesk: Look up support tickets, customer issues, and SLA status. - Trello: Check boards, cards, and list progress. - PayPal: Check transactions, process refunds, and manage payments. ## API Quick Start: Zero to Meeting Summary The full meeting lifecycle via API. Requires a Pro plan and an API key. ### Step 1: Get an API key Create one at https://vernix.app/dashboard/settings. All requests use: Authorization: Bearer kk_your_api_key_here ### Step 2: Create a meeting POST https://vernix.app/api/v1/meetings Content-Type: application/json { "title": "Weekly Standup", "joinLink": "https://meet.google.com/abc-defg-hij" } Response (201): { "data": { "id": "meeting-uuid", "title": "Weekly Standup", "status": "pending", ... } } Optional fields: "agenda" (string, up to 10k chars), "silent" (boolean, text-only mode), "noRecording" (boolean), "autoJoin" (boolean, skips step 3). ### Step 3: Join the agent to the call POST https://vernix.app/api/v1/meetings/{id}/join Response (200): { "data": { "botId": "recall-bot-id", "status": "active" } } The agent joins the video call. Status transitions: pending -> joining -> active. In voice mode (default), the agent listens and responds to questions when you say "Vernix". In silent mode, it responds via the meeting chat when someone types @Vernix. Shortcut: Use "autoJoin": true in Step 2 to create + join in one call. ### Step 4: Wait for the call to happen While the meeting is active, the agent: - Transcribes the conversation in real time - Responds to questions using RAG (voice or chat) - Connects to your configured integrations (Slack, Linear, GitHub, etc.) You can check the meeting status at any time: GET https://vernix.app/api/v1/meetings/{id} The "status" field will be "active" while the call is ongoing. ### Step 5: Stop the agent and trigger processing POST https://vernix.app/api/v1/meetings/{id}/stop Response (200): { "data": { "status": "processing" } } This stops the agent, then automatically: 1. Generates an AI summary of the meeting 2. Extracts action items / tasks from the conversation 3. Sets status to "completed" when done ### Step 6: Get the results Once status is "completed", retrieve the outputs: Meeting details (includes summary in metadata): GET https://vernix.app/api/v1/meetings/{id} Full transcript: GET https://vernix.app/api/v1/meetings/{id}/transcript Extracted tasks: GET https://vernix.app/api/v1/meetings/{id}/tasks ### Step 7: Search across meetings Search all your meeting transcripts and knowledge base: GET https://vernix.app/api/v1/search?q=what+did+we+decide+about+pricing ### Meeting Status Lifecycle pending -> joining -> active -> processing -> completed -> failed (if something went wrong) - "pending": Created, agent not yet joined - "joining": Agent is connecting to the call - "active": Agent is in the call, transcribing - "processing": Call ended, generating summary and tasks - "completed": Summary and tasks ready - "failed": Something went wrong (can retry join) ### One-Command Example (curl) Create a meeting and join immediately: curl -X POST https://vernix.app/api/v1/meetings \ -H "Authorization: Bearer kk_your_api_key" \ -H "Content-Type: application/json" \ -d '{"title": "Quick Sync", "joinLink": "https://meet.google.com/abc-defg-hij", "autoJoin": true}' Stop the meeting and get the summary: curl -X POST https://vernix.app/api/v1/meetings/{id}/stop \ -H "Authorization: Bearer kk_your_api_key" curl https://vernix.app/api/v1/meetings/{id} \ -H "Authorization: Bearer kk_your_api_key" ## API Reference - API Documentation (interactive): https://vernix.app/docs - OpenAPI Spec (machine-readable): https://vernix.app/api/v1/openapi.json - Authentication: Bearer token with API key - Rate Limits: 60 requests/minute standard, 10 requests/minute for search/agent operations - Daily Quota: 1000 requests/day (Pro plan) - Response Format: { "data": ..., "meta": { "hasMore": bool, "nextCursor": string } } - Error Format: { "error": { "code": "NOT_FOUND", "message": "Meeting not found" } } - Pagination: Cursor-based. Pass ?limit=20&cursor=opaque_string ### All REST API Endpoints Meetings: - POST /api/v1/meetings — Create meeting (+ optional autoJoin) - GET /api/v1/meetings — List meetings (?status=completed&limit=20&cursor=...) - GET /api/v1/meetings/:id — Get meeting details + summary - PATCH /api/v1/meetings/:id — Update meeting (title, joinLink, agenda, silent, muted) - DELETE /api/v1/meetings/:id — Delete meeting and all associated data - POST /api/v1/meetings/:id/join — Join agent to call - POST /api/v1/meetings/:id/stop — Stop agent, trigger summary + task extraction - GET /api/v1/meetings/:id/transcript — Get full transcript with speaker labels - GET /api/v1/meetings/:id/tasks — List tasks for this meeting - POST /api/v1/meetings/:id/tasks — Create a task manually Tasks: - GET /api/v1/tasks — List all tasks across meetings (?status=open) - GET /api/v1/tasks/:id — Get task details - PATCH /api/v1/tasks/:id — Update task (title, assignee, status, dueDate) Search: - GET /api/v1/search — Semantic search (?q=query&meetingId=...&limit=10) Integrations: - GET /api/v1/integrations — List connected integrations (Slack, Linear, GitHub, etc.) Knowledge Base: - GET /api/v1/knowledge — List documents (?meetingId=...) - POST /api/v1/knowledge — Upload document (multipart/form-data, field: "file") - GET /api/v1/knowledge/:id — Get document details + download URL - DELETE /api/v1/knowledge/:id — Delete document ### MCP Server For AI assistants (Claude Desktop, Cursor, etc.) — same API key, richer tool interface. MCP Endpoint: https://vernix.app/api/mcp Transport: Streamable HTTP (with SSE fallback) Claude Desktop config: { "mcpServers": { "vernix": { "url": "https://vernix.app/api/mcp", "headers": { "Authorization": "Bearer kk_your_api_key" } } } } Available MCP tools: - search_meetings — Search transcripts and knowledge base via vector similarity - list_meetings — List meetings with optional status filter - get_meeting — Get meeting details including summary and agenda - get_transcript — Get full transcript with speaker labels - list_tasks — List action items across meetings - create_task — Create a task for a specific meeting - vernix_join_call — Create a meeting and join the agent to a call - vernix_stop_call — Stop the agent and trigger processing - vernix_search_meetings — Semantic search with structured results - vernix_search_tasks — Search and filter tasks - list_integrations — List connected integrations (Slack, Linear, GitHub, etc.) ## Links - Homepage: https://vernix.app - Tool Integrations: https://vernix.app/features/integrations - Meeting Memory: https://vernix.app/features/meeting-memory - Knowledge Base: https://vernix.app/features/context - Pricing: https://vernix.app/pricing - FAQ: https://vernix.app/faq - Contact: https://vernix.app/contact - API Docs: https://vernix.app/docs - OpenAPI Spec: https://vernix.app/api/v1/openapi.json