Mastermatch was an AI match making progressive web application aimed to connect professionals based on a passive AI algorithm which suggests professionals to match based on their information.
Mastermatch was a startup which tested out their application in 2023-2024. We inherited a good chunk of infrastructure and architecture, but not a whole lot of functional secure code, documentation, setup instructions, automated tests, or a version control structure.
Despite all of this, we took on the project. The app needed to be feature complete, fully styled, maintainable, documented, and have its main selling point developed: AI matchmaking.
At its peak Mastermatch had a few hundred users but got taken offline due to hosting costs exceeding investments.
For this project we took on the full stack of technical responsibilities among which:
- Documenting all code so it's easier to maintain and hand over
- Ensure all functionalities are securely implemented without vulnerabilities
- Expand and maintain application features
- Expand and maintain the web app's frontend
- Develop the AI matchmaking algorithm
- Develop, improve, and maintain data structures in the database
During the weekly "Match Time" event, users could tell the AI what kind of matches they were interested in, or - if the AI had enough information - we could fill this in for them automatically.
During the weekly "Match Time" event, users could tell the AI who they are and what they have to offer, or - if the AI had enough information - we could fill this in for them automatically.
After filling in who you are and what you're looking for, the AI would automatically suggest the best possible matches for you and suggest you to others.
Besides AI matchmaking, the Mastermatch platform provided events, networking circles, company circles, messaging boards, and a match request feed to allow users to also manually look for the next business match.
Every two weeks we did a sprint with its own deliverables. The deliverables were defined by the product owner each sprint. After which these were pushed to our staging environment. If everything worked well in staging, the deliverables were also pushed to production.
Some of the technologies used in this project were:
- ASP dotnet for backend
- Identity Server for authentication
- React with material UI for frontend
- Mollie for payments and subscription
- MongoDB for the database
- AWS EC2 for server hosting
- AWS S3 for user file hosting
- LamaSharp library & Vicuna7b LLM for AI functionalities
- Windows Server with IIS for server management
- Azure DevOps for documentation and version control
- Selenium for automated tests
