Projects
Impact
About Klub
Klub Capital India is a revenue based financing platform, that helps digital businesses and SMEs grow by offering flexible funding.
Klub partners with NBFCs, HNIs & financial institutions to provide capital.
Whether a business is in its early, growth, or late stages, Klub offers tailored funding solutions for recurring expenses like marketing, inventory, and capex, helping businesses thrive at every stage.
650+ brands have raised funding with Klub.
How it all began?
The lender platform at Klub had already implemented smoother workflows and automation, but support teams still handled numerous repetitive queries, particularly around loan statuses and document requests. This led to high support ticket volumes and occasional delays in response times, impacting overall efficiency.
Setting the direction
Since the AI agent was a new feature, our product and development teams had to learn about language learning models from scratch. We adopted a Minimum Viable Product (MVP) approach to keep the initial functionalities simple for Phase 1, focusing on automating routine queries, providing real-time assistance, and enhancing lender productivity while ensuring room for future enhancements.
For simplicity sake, we called it KAM or Key Account Manager.
Simplifying the Complex
Before jumping into designing the AI agent, it was essential to identify its exact need. I collaborated with our marketing and customer support teams, who managed calls from interested brands and lenders. This helped me understand recurring questions, pain points, and areas where users needed more assistance.
Since I was responsible for designing the lender side (CPP) platform, I focused on lender queries related to task updates, document verification, and loan status checks. On the brand side, common questions revolved around onboarding, required documentation, and platform navigation.
Additionally, I reviewed customer support scripts to analyze how conversations were humanized. This ensured that interactions with the AI agent felt natural and user-friendly, minimizing mechanical responses while providing accurate assistance.
Workflow
The development team worked out the mechanism for the functioning of the AI agent.
The process begins with the user submitting a loan application, which triggers document parsing that extracts relevant details from the uploaded files.
The system then conducts eligibility checks using GST documents and the last three months' bank statements, which help assess the financial health of the borrower. The results from the eligibility check are then passed to a rule-based engine, which applies predefined lending criteria to determine if the applicant qualifies.
For a seamless interaction, the platform allows users to manage tasks effortlessly with voice commands, ensuring a hands-free experience for relationship managers. Once the eligibility data is processed, it undergoes real-time validation, ensuring the decision-making is quick and accurate.
Finally, the credit decision is based on the eligibility analysis and is communicated to the borrower through user-friendly notifications. These notifications are designed to be clear and actionable, providing a smooth, interactive experience for the user.
Designing the solution
Once the queries and needs were clear, the next step was to design the AI agent’s user interface. Our design team explored multiple iterations, drawing inspiration from Siri and other LLM-based assistants. We finalized a version tailored for the Brands App, with slight modifications for the Lender Platform.
The next challenge was scripting and building seamless voice-activated interactions. Using ProtoPie, I designed simple conversation flows, ensuring all relevant pages were connected for smooth navigation.
The design prioritized lenders’ needs: offering an overview of key information and automating loan application tasks within the CPP, reducing the need for human intervention while maintaining efficiency.
To conclude
KAM or the ai agent successfully addressed key pain points for lenders by providing instant assistance, automating routine tasks, and improving operational efficiency. This feature not only enhanced user experience but also optimized internal workflows, demonstrating how thoughtful design can drive tangible business outcomes. As the AI agent continues to evolve, it lays the foundation for future innovation and scalability within the lender platform.
