Loading...
Back to case studies
Live dashboard
Voice of the Customer analytics platform
An analytics pipeline for insurance call centre transcripts, extracting structured customer and operational insight for sentiment, churn risk, topic classification and agent performance.
Workflow evidence
Screenshots from the systems and workflows.
These visuals come from the project portfolio and show the workflow, dashboard or training assets behind the case study.
Problem
The operational problem behind the work.
Call transcripts contain valuable customer signals, but manual review does not scale. The goal was to turn unstructured conversations into useful operational insight without losing the nuance of the customer journey.
Approach
How the system was shaped.
01
Designed an end-to-end transcript processing workflow.
02
Extracted structured fields from each call, including sentiment journey, churn risk, topic classification and agent performance signals.
03
Served outputs into a dashboard for review and exploration.
04
Created a video walkthrough and live dashboard pathway to demonstrate the system.
Architecture
Components that made the work production-minded.
Transcript ingestion workflow
Gemini 2.5 Flash extraction layer
ArangoDB storage for structured call outputs
n8n orchestration
Docker and Traefik deployment
Chart.js dashboard served through a webhook
Outcomes
What changed or became possible.
16 structured fields extracted per call
Sentiment journey and churn risk signals generated from transcripts
Dashboard served via webhook
Video walkthrough available for demonstration
Responsible AI
Why governance is part of the implementation.
Analytics supports service improvement and review rather than replacing human judgement.
Structured outputs make model behaviour easier to inspect than free-form summaries alone.
Customer data workflows should be governed with privacy-aware access and retention decisions.
Stack
Tools and platforms used.
n8nGemini 2.5 FlashArangoDBDockerTraefikChart.js
Similar challenge?
Start by assessing the workflow, data and risk.
The AI Readiness Audit is the cleanest first step when you need to decide what to automate, what to avoid, and what should be governed before a production build.