PrediCX - From ML Platform to sundry Activelearning Classifier
João Ribeiro
Like this project
Posted Dec 4, 2024
REST API driven ML engine that pivoted into a B2B SaaS product for automatic labeling of customer date through customer-trained or pre-trained models.
Likes
0
Views
0
Clients
Warwick Analytics
Implemented a distributed and dynamically scalable ML engine in the AWS cloud to provide support for running turn-key ML algorithms (R-case, Clustering and Classification) on large tabular datasets.
- Backed by a public facing REST API and an internal event driven architecture.
Contributed to the pivot of the engine into a B2B-focused SaaS product, PrediCX.
Supporting the automatic classification of customer data based on Active learning technique and featuring a market of pre-trained models covering a wide range of industries.
- Implemented the front-end SPA application for the product
- Designed and implemented the training and classification pipelines backing the product through a dynamically, and cost-effective, scalable architecture.
Techstack: Java, Python, Apache Spark, Amazon Aurora (MySQL), Angular, Docker, Terraform, Packer, AWS.