1. Server events
Flireo AI
  • Introducing Flireo AI
    • Introduction
    • Core concepts
    • Getting started
    • API overview
    • Errors
  • Assistants
    • Phone numbers linking
  • Phone numbers
    • Overview
    • Available numbers
    • List phone numbers
    • Order a phone number
    • Phone number orders
  • Verification groups
    • Overview
    • Available numbers
    • Create verification group
    • List verification groups
    • Order number via verification group
  • Server events
    • Assistant types
    • Assistant-request Event
    • Tool-calls Event
    • End-of-call-report Event
    • Status-update Event
    • Control live calls
API ReferenceDocumentatie
API ReferenceDocumentatie
Linkedin
  1. Server events

End-of-call-report Event

Every time a conversation ends, Flireo sends an end-of-call-report event to your external server. This event gives you a complete summary and analysis of the call—ready for automation, email notifications, or database logging.

Example Payload#

Here’s what you might receive:
{
  "eventTimestamp": 1756294820944,
  "eventType": "end-of-call-report",
  "callId": "cc1abf4a-d005-44a8-825f-702334e8d4c8",
  "fromCaller": "+31623731537",
  "toAssistant": "+31850835003",
  "transcriptText": "AI: Goedemiddag, met Julia van Voorbeeld Bedrijf, hoe kan ik u vandaag helpen?\nGebruiker: Hallo Julia, ik heb een vraag over jullie diensten.\nAI: Natuurlijk! Wij bieden oplossingen in de techniek, bouw en logistiek. Kunt u iets meer vertellen over waar u naar op zoek bent?\nGebruiker: Ik ben geïnteresseerd in technische functies. Kunt u daar meer over vertellen?\nAI: Zeker! Wij zijn gespecialiseerd in het verbinden van technische professionals met bedrijven die innovatie waarderen. Wat voor technische ervaring heeft u?\nGebruiker: Ik heb voornamelijk in softwareontwikkeling gewerkt.\nAI: Dat klinkt goed! Heeft u een voorkeur voor frontend, backend of full-stack rollen?\nGebruiker: Vooral backend.\nAI: Perfect, we hebben meerdere backend functies beschikbaar. Wilt u dat ik u meer informatie stuur of een afspraak inplan?",
  "summaryText": "Dit gesprek was een professionele introductie tussen Julia (AI) en een potentiële kandidaat die interesse heeft in technische functies. De AI gaf informatie over de diensten van Voorbeeld Bedrijf, met focus op technische mogelijkheden. De gebruiker toonde interesse in backend ontwikkeling en de AI bood aan meer details te sturen of een afspraak te maken. Het gesprek was informatief, vriendelijk en leidde tot een duidelijke vervolgstap.",
  "recordings": {
    "linkStereo": "https://vapqsrlwmlidfqoenpcw.supabase.co/storage/v1/object/sign/recordings/example.mp3?token=EXAMPLE_TOKEN"
  },
  "callLengthSeconds": 132.5,
  "userMetadata": {
    "candidateId": "cand_123456",
    "preferredRole": "backend"
  },
  "analysis": {
    "summary": "Julia (AI) gaf een helder overzicht van de technische functies bij Voorbeeld Bedrijf. De gebruiker was betrokken en deelde relevante ervaring. Het gesprek had een positieve uitkomst met een duidelijke vervolgactie.",
    "structuredData": {
      "sentiment": 9,
      "ai_ervaring": 10,
      "hoofdonderwerp": "Gebruiker informeert naar technische carrièremogelijkheden.",
      "vraag_beantwoord": true,
      "gesprek_kwaliteit": "Uitstekend",
      "bedrijf_waardering": 9,
      "belangrijkste_punt": "Gebruiker is geïnteresseerd in backend rollen en vraagt om meer informatie.",
      "vervolgactie_nodig": true,
      "gespreksduur_gevoel": "Precies goed",
      "technische_problemen": false,
      "nieuw_veld_11": "N.V.T."
    },
    "successEvaluation": "true"
  }
}

What’s in the Report?#

Basic Call Data:
callId, fromCaller, toAssistant
eventTimestamp
callLengthSeconds (duration)
recordings (link to stereo recording file)
Any userMetadata you specified in the assistant-request
Transcript:
Full transcript of the conversation—every spoken word from AI and user.
Summary:
Natural language summary generated by your assistant, outlining the main topic, important points, and any agreements or next steps.
Analysis:
Custom analysis report, structured exactly as defined in your assistant configuration (see /assistants/[id]/analysis).
structuredData: Key metrics and fields (e.g., sentiment score, issue resolved, action needed, technical issues).
successEvaluation: Did the call meet your success criteria?
All fields are customizable via the dashboard.

How to Use the Data#

Automatic Email Notifications:
Format the summary and analysis into a rich HTML email for instant review by your team.
Database Logging:
Store every report in your CRM or analytics database. Track metrics, build dashboards, and improve your processes.
Workflow Automation:
Trigger follow-up actions (e.g., schedule a meeting, send documents, escalate support issues) based on analysis results.
Quality Control:
Use the transcript and structured analysis to monitor call quality, evaluate agents, and optimize your customer experience.

Customizing Your Analysis#

You can fully control which fields are sent in the analysis object.
Go to /assistants/[id]/analysis in the dashboard to build your own schema—track anything from sentiment, intent, technical issues, satisfaction, to custom business metrics.
These fields appear in every end-of-call report, enabling deep integration and automation for your business.

Important Notes#

No response required:
You do NOT need to respond to the end-of-call-report event. It’s sent for you to use as you wish.
The report is sent immediately after a call ends.
All data is available for further automation, reporting, and compliance.

Next:
Explore how to customize your analysis schema or discover best practices for automating post-call actions.
Modified at 2025-08-28 08:18:26
Previous
Tool-calls Event
Next
Status-update Event
Built with