Simplifying Transactions for Education Financing

A tech solution that easily allows teachers to get the funds they need for their classrooms. Learn more about our solution for ClassWallet.

Big Data
Machine Learning
LLMs
Simplifying Transactions for Education Financing

About ClassWallet

ClassWallet is a way to simplify teacher funding; a tech solution that easily allows teachers to get the funds they need for their classrooms. ClassWallet serves over 6,200 schools, 20% of all US State Agencies, and over 4.1 million students.

Their digital wallet solution is made for both institutions and individual users. For organizations like schools and grant providers, ClassWallet streamlines the distribution of funds, such as scholarships or grants. It automates the disbursement process, reducing administrative burden and providing centralized records for easy tracking and reporting.

On the other hand, ClassWallet empowers families participating in programs like Educational Savings Accounts (ESAs). It functions as a digital wallet for parents, allowing them to manage their ESA funds, make direct tuition payments to qualified schools, and purchase approved educational materials from a dedicated marketplace within the platform.

The Challenge

The platform needed improvements in different areas: better reimbursement management, sorting out and categorizing different items from suppliers, and gathering information from pictures of receipts and invoices (store names, dates, total costs, and details).

Until we started collaborating, every process was done manually by ClassWallet. 

The Solution

To provide a complete solution for ClassWallet, we decided to divide their needs in a two-phase roadmap. 

In Phase One, we created an automated classification system that categorized each item using the United Nations Standard Products and Services Code. Using the most granular level within the five-tiered hierarchy, we employed a sequence text classification model trained on ClassWallet's labeled dataset, which contained millions of preprocessed data points.

In Phase Two, we implemented Optical Character Recognition (OCR) to extract information from invoice images automatically. These images are also turned into files that feed the classification system.

Impact

Ready to jump on the next big thing?

Unlock the power of AI and Machine Learning with our cutting-edge solutions.