Inspiration
Managing transaction data can quickly become overwhelming for small businesses. The data is there, but turning it into useful decisions often requires manually going through transactions, customers, amounts, and patterns.
That inspired us to build Salah — an AI-powered financial intelligence tool that turns transaction data into clear, actionable business insights.
For this hackathon, we took that idea further by integrating the product with the PayPal developer platform, making PayPal transaction data a core part of the experience rather than simply adding PayPal as a payment option.
What We Built
Salah brings transaction and customer data into one place and uses AI to help understand what is happening in the business.
Instead of manually analyzing hundreds or thousands of transactions, a user can get insights around spending patterns, customer behavior, transaction trends, and potential areas that deserve attention.
The goal is simple: turn raw financial data into decisions that are easier to understand and act on.
How We Built It
We started with the transaction-analysis workflow from our existing prototype and significantly evolved it for this hackathon around the PayPal ecosystem.
The application connects transaction data with an AI analysis layer and presents the resulting insights through a simple dashboard. We focused on keeping the experience understandable even for users who are not financial analysts.
The system was tested against a dataset containing 1,874 transactions from 152 customers, with 67 tests passing, helping us validate the core transaction-analysis workflow.
What We Learned
One of the biggest lessons was that adding AI is not enough to make a product useful. The real challenge is deciding what information the AI should extract, how it should be presented, and how it can help someone make a better decision.
We also learned how important it is to make a third-party payment platform part of the actual product workflow. Working with PayPal pushed us to think beyond a simple payment integration and design around transaction intelligence.
Challenges
The biggest challenge was making the financial insights useful without overwhelming the user with raw data.
We had to think carefully about how to structure transaction information, identify meaningful patterns, and turn those patterns into concise insights. Integrating the PayPal workflow while keeping the existing product architecture stable was another challenge.
The hackathon also pushed us to balance three things at once: AI capabilities, reliable financial data, and a simple user experience.
Salah is our attempt to bridge that gap — taking transaction data that already exists and turning it into something a business can actually understand and use.
Log in or sign up for Devpost to join the conversation.