Overview
Contact center managers at Dialpad needed a way to measure agent performance without spending hours manually reviewing calls. I built Digital Scorecards to automate this. Real-time performance tracking with customizable scoring criteria.
I designed and built the full-stack solution as a Software Engineering Intern. From architecting the scoring system to shipping to production. The feature now serves 1000+ agents and reduced manual review time by 60%.
The Problem
Traditional performance reviews in contact centers were broken. Managers spent hours listening to call recordings and filling out subjective evaluation forms. Agents received delayed, inconsistent feedback. At scale, this couldn't work.
Dialpad needed an automated system that could score interactions based on configurable criteria. Give managers instant insights and give agents clear feedback on their performance.
Technical Implementation
Built the frontend with Vue.js. Real-time dashboard updated via WebSocket connections. Managers can define custom evaluation criteria with weighted scoring, drill down into individual agent performance or view team-level trends.
The Django backend handles score calculation, aggregation, and automated report generation. Deployed on Google Cloud Platform with PostgreSQL and Memcached. The system handles high query volumes with sub-200ms response times and 99.9% uptime.