GenAI-Powered Negotiation Scripts Lift Collections by 22%
Generative AI-driven dynamic scripting for a private sector bank's collections unit — personalising negotiation conversations at scale.
The Challenge
A private sector bank's collections centre of 120 agents was relying on static call scripts that failed to account for individual customer circumstances — income shocks, medical events, job changes — resulting in collections interactions that felt scripted and adversarial, driving high call refusal rates.
Our Approach
Deployed a GenAI scripting engine trained on 18 months of successful collections conversations and customer profile data.
Engine generates real-time prompts during live calls — surfacing the most likely resolution pathway based on customer sentiment, payment history, and stated reason for default.
Integrated with the bank's CRM via API — agent sees prompts on a side-panel without switching screens.
Ran a 6-week A/B test: 60 agents on GenAI-assisted scripts vs 60 on standard scripts. GenAI cohort outperformed on every metric.
Continuous model fine-tuning: weekly feedback loop from QA team flagging high and low-quality interactions.
Related Engagements
Ready to Achieve Results Like These?
Tell us about your operational challenge — we'll show you exactly how we'd approach it.