No-Code Data Science Platform
The no-code data science platform allowed anyone, regardless of data science expertise, to connect their datasets wherever they lived, visualize them, analyze them, and create A.I. models off of them.
Prior to building Apteo's ecommerce personalization platform, my startup, Apteo, created a no-code data science tool that could be used by anyone. The goal of the product was to provide every employee in a company with the tools they needed to make data-driven business decisions without having to get trained in data science. The tool had data connectors that would allow anyone to connect data from a variety of different data sources (spreadsheets, databases and data warehouses, API tools, etc) and centralize that in a single place. From there, they could create and serve their own A.I. models on this data, create visualizations, run correlations, run a variety of analytics on that data.
Key features of the project included:
- The ability to connect datasets from multiple sources, including databases, data warehouses, files, and third party platforms
- The ability to visualize and analyze data
- The ability to build an A.I. model to predict any column in any dataset using a point-and-click UX, along with the ability to identify key factors within that model and the ability to serve that model via an API
The platform was used by a variety of businesses for many different use cases, including predicting customer churn, running correlations on alternative data for finance, and creating dashboards and visualizations for reporting. The product was eventually turned into Apteo's personalized ecommerce marketing platform.