Mental Health MLOps Project - Stress Prediction Using Multimodal Inputs
In this project, I have implemented the end-to-end MLOps life cycle, including infrastructure provisioning using Terraform, modeling and experiment tracking using MLflow, orchestration with Mage, deployment using Flask, Lambda and Kinesis, monitoring with Evidently and Grafana, and best practices such as integration tests with Localstark, unit testing, linting and formatting, Makefile and pre-commit hooks.
MLOps project pipeline