Software Engineer Intern
KPMG
Worked on a multi-agent AI system to support call centre operators in real time.
Key responsibilities
- Designed a real-time multi-agent AI system for telecom call centres, covering the full support lifecycle from live transcription to automated service order creation.
- Built a Python/LangGraph orchestration backend coordinating specialised agents for ASR, sentiment analysis, contact-reason classification, and remote diagnostics.
- Integrated CRM, network monitoring, and interaction history into a structured live summary surfaced to the human agent during the call.
- Implemented automated service order pre-filling with SLA context and appointment scheduling, triggered when remote resolution was not viable.