2024
FinanceOps AI
financeops.ai
Background
This case study explores how our AI-powered VoIP calls have transformed collections for a fintech startup, making high-volume outreach more efficient, easier to manage, and improving overall recovery rates.
Designing a Solution
The initial iterations were made via back and forth with PM requirements and sales calls. The PRD contained specs that were as follows:
Pre-Call Specs : Agents previously lacked a centralized view of debtor info and context, making it difficult to prioritize accounts and strategize calls effectively. Pre-call AI guidance should solve this by preparing agents with actionable insights. Objective: Prepare the agent with insights and context to improve call effectiveness.
During Call Specs: Agents faced difficulty managing conversations in real-time, understanding debtor sentiment, and accurately capturing outcomes. Live AI support should ensure consistent, informed, and productive calls. Objective: Provide real-time guidance and support to maximize collections efficiency.
Post Call Specs: Previously, follow-ups and performance tracking were manual and error-prone, leading to inconsistencies. Post-call AI features should allow continuous improvement, accurate tracking, and better-informed future outreach. Objective: Capture insights, update account information, and optimize future interactions.
The Design Process
In this phase, we synthesized insights from all sales calls with top fintech, utility, and healthcare companies and combined them with the Product Requirement Document (PRD) analysis using ChatGPT. The goal was to systematically identify pain points with the highest business impact to prioritize solutions in the upcoming product design.
Using AI-assisted analysis on the PRD, we were able to quantify the severity and frequency of these pain points, highlighting which hallenges had the most significant effect.
Results and Impact
High Adoption Among Clients:
The VOIP/Voice AI feature achieved a 72% adoption rate within 6 months, outperforming other channels like SMS and email. It played a key role in closing deals with Shell and TD Bank.
Substantial Contribution to Collections:
Over 40% of collections on the platform are conducted via Voice AI/VOIP, making it the 4th most utilized feature across all offerings.Significant Revenue Recovery:
The feature helped recover $28 million in total, spanning 25,000 debtors in the US.Enterprise-Level Usage:
Currently used by four companies, each with a team strength of 15 agents, demonstrating scalability for mid-sized collections teams.Time Saved on Repetitive Tasks:
Automation through VOIP/Voice AI reduced time spent on repetitive calling tasks by 35%, equivalent to approximately 45 man-hours per agent per week, allowing teams to focus on high-value, complex accounts.
Read Full Case Study