saina project

AI-Enhanced Fraud Detection: Enhances the security of Square's payment ecosystem and reduces chargebacks for sellers,

Comment

Inspiration

We are big supporters of SDGs and this project inspired us because it is an enabler for Sustainable Development Goal 9 - Trade, Industry & Infrastructure - Build resilient infrastructure, promote sustainable industrialization and foster innovation, we cann play our part and provide affordable and reliable logistics services for SMEs in Africa because logistics account for more than 70% of a product price, this solution eliminates middlemen , fraudsters and increase profit for farmers, Artisans, Creators and SMEs.

The other inspiration for building saina lies in the fusion of innovation because it's the innovative approach that sets the solution apart, the solution is in the intersection of AI, Mobile and Cloud. Cutting-Edge Technology we have used like Google cloud is at the forefront of AI and machine learning technologies. Leveraging Google AI brings the latest advancements in machine learning, which is exciting for the developers and the users of saina.

What it does

Saina is fraud detection system that leverages Google AI to analyze transaction data and detect potential fraudulent activities in real-time. This enhances the security of Square's payment ecosystem and reduces chargebacks for sellers, it offers a comprehensive and innovative solution to tackle cybersecurity threats and challenges. Leveraging Google Cloud's advanced capabilities, it provides real-time threat detection, predictive analysis, and automated incident response, setting it apart as a unique and effective cybersecurity solution.

The reason Saina should be adapted are, it provides

  • Real-time threat detection and proactive response.
  • Scalable and adaptable to organizations of all sizes.
  • Integration of external threat feeds for comprehensive threat intelligence.
  • Predictive analysis for future threat identification.
  • Automation of incident response for faster mitigation.
  • Compliance reporting and evidence collection. ## How we built it Data is key, No AI projects beigins without data, so we gathered transaction data from Square's payment ecosystem. This includes transaction details, customer information, and any other relevant information, this integrated data sources into the saina platform, making it accessible for analysis.

Basically, we implement a Secure Payment Flow in our fullstack,so that when a user proceeds to checkout, send a request to the server which is a VM in GCP to initiate a payment. This are the logs we are interested in in, we actually use streaming to get real-time data warehouses and then create a dashboard of real-time information and notifications incase a fradulent activity is predictde by our AI model in real time.Mobile Application: Used React-native to build a mobile apps that works well on both Android and iOS,the Back-end is built using Nodejs, MongoDB and Express framework.

GCP, Where the fun begins

Alt text

Challenges we ran into

My teammate Christine and I are fully employed and finding time to study and take online courses to build the project was tough and really challenging.

Other challenges we ran into from a data engineering standpoint, was the way to constructed the pipeline to minimize latency at every single step. If it's a Dataflow job, we designed it so that as many elements as possible are happening in parallel because if data comes in late, especially when it comes to fraud detection, it's no longer valuable, especially, during an emergency .

At the end we had to deal with challenges associated with streaming applications, you're talking about the 3V's;

volume challenge because the data never stops coming and quickly grows, then velocity and It is important to design systems that can handle that extra load. Variety of data is the third challenge. If you are using only structured data, data coming from a mobile app, that is easy enough to handle. But what if you have unstructured data like voice data or images?

Accomplishments that we're proud of

We learned streaming in the Cloud can help us here, so we took a Data Engineering, Big Data, and Machine Learning on Google Cloud Plartform Specialization on Coursera, passed all requirements and earned course certificates.

What we learned

During the course work, on the volume side we found a tool to assist in autoscaling processing and analysis, so that the system can handle the volume. On the velocity side, a tool that can handle the variability of the streaming process was available. On the variety side, we looked at how artificial intelligence could help us with unstructured data. ## What's next for kafunge Saina to be adapted by square's largest sellers as their one-stop fraud detection intelligence platform that leverages Google Cloud's advanced features to proactively detect and mitigate cybersecurity threats in real-time.

We intent to stay agile and adapt to the evolving cybersecurity landscape to remain competitive and meet customer needs.

Build a sustainability and profitability business requiring a combination of smart pricing, diversification of revenue streams, excellent customer service, and a commitment to ongoing improvement and innovation.

Share this project:

Updates