Leading the design of an autonomous AI video agent capable of conducting verification calls for financial institutions, transforming a manually intensive process into an automated, scalable solution while maintaining compliance and security standards.
Role
Product Designer
Industry
FinTech / Video Technology
Duration
Ongoing
Team Size
Cross-company collaboration
Status
MVP in Development
Context
During my tenure working on video technology, we identified a critical inefficiency in video assistance: human agents spent only about 40 minutes of their 8-hour workday in actual calls, resulting in significant resource underutilization. As AI capabilities expanded, we saw an opportunity to revolutionize video assistance through autonomous agents.
The financial sector, particularly loan verification, presented a perfect initial use case. Banks were already using our platform to record loan confirmations for legal protection, making it an ideal testing ground for AI automation while maintaining the essential recording and verification aspects.
Opportunity
Analysis revealed that company representatives were spending nearly 7 hours daily in non-call activities, creating massive inefficiencies in cost structure. Additionally, the repetitive nature of loan verification calls made them ideal for automation:
Standardized questions
Predictable conversation flows
Clear success metrics
High volume of similar interactions
Significant cost-saving potential
Solution
Developed a comprehensive AI video agent system through collaboration with a major software development house:
Created detailed conversation flow architecture
Designed user interaction patterns
Developed failsafe mechanisms
Established compliance checkpoints
Built scalable agent customization framework
The MVP focuses on loan verification, with the agent conducting structured interviews while maintaining recording capabilities for legal compliance. Future iterations will expand to:
Student assistance
Medical consultations
Customer service
Customizable agent personalities
Results
Created working prototype of AI video assistance agent
Established strategic partnership with major software development company
Projected 60% cost reduction for clients
Developed comprehensive conversation flow architecture
Successfully demonstrated MVP concept for loan verification
Key Takeaways
Strategic partnerships can overcome technical limitations
Start with highly structured use cases for AI implementation
Design for both efficiency and human comfort
Maintain compliance while innovating
Balance automation with human oversight
