AgrIntel: Gen AI Assistant for
Agri Bank
A small project for - Generative AI-powered responsive web app for agricultural banking documentation.
Client: Capital Farm Credit
Duration: 2 months.
Category: Fintech, AI
Role: UX Research, UI Design, Product Strategy
_edited.png)
_edited.png)
Overview
Agricultural bankers deal with an enormous volume of internal documentation, including farmer loan histories, regional policies, compliance guidelines, and association records.
This information is often stored in fragmented PDFs and legacy systems, making access slow and error-prone.
Challenge
Agricultural bankers navigate a complex web of internal documentation—ranging from fragmented PDFs to outdated legacy systems—to access critical information like farmer loan histories, compliance policies, and regional guidelines.
This scattered ecosystem leads to frequent delays, human error, and inefficiencies in decision-making, directly impacting the speed and accuracy of loan processing and regulatory compliance.
Research
Stakeholder Interviews
-
Who: Bankers, IT admins, loan officers, and support staff etc.
-
Goal: Understand current pain points, workflows and expectations.
Key Insights:
-
Users frequently referenced outdated PDFs or branches for help.
-
Time to locate specific policy or borrower information avg.15–25 mins.
Competitive Benchmarking:
-
Referenced platforms: ChatGPT for conversational interaction; Microsoft Copilot for secure enterprise integrations.
-
Noted a lack of industry-specific generative AI tools in agri-finance.
Banker
Farmer
Admin
Insights
User Personas
-
The Banker – Needs quick, accurate answers to policy or questions.
-
The Farmer – Wants simple explanations and guidance on loan status
-
The Admin – Manages data integrity, user roles, and AI learning loops.
Goals
-
Create a frictionless experience for asking questions.
-
Build trust through clarity, accuracy, and consistent UI feedback.
.png)
Design & Wireframing
We created a centralized dashboard that organizes key data—farmer profiles, loan history, and compliance guidelines—into an intuitive, searchable interface.
Using low-fidelity wireframes, we prioritized quick access and contextual relevance. Filters, smart search, and collapsible content blocks were introduced to streamline navigation. After usability testing, we refined the layout for clarity, speed, and reduced cognitive load.
.png)
Result
Looking ahead, there are several opportunities to enhance Agri Intel’s capabilities and reach. Integration with CRM systems would allow users to take direct action—such as initiating loan updates or follow-ups—based on the AI’s responses.
Lastly, a mobile-first version with offline caching could further extend accessibility to rural areas with limited internet connectivity.
WELCOME WELCOME MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK MORE WORK WELCOME
