Ecosyste.ms sponsors
An open API service aggregating public data about GitHub Sponsors.
An open API service aggregating public data about GitHub Sponsors.
ML Architect. Father of two daughters.
Funding Links: https://github.com/sponsors/arnabbiswas1
I am a Machine Learning Developer from Bangalore, India. Around 16 years back, I started my career as a software developer. Over years, I have worked for multiple organizations in the Telecom & Networking domain, e.g. , Cisco Systems, Nokia Siemens Networks etc.
As an ML architect, my current area of focus is Machine Learning for Predictive Maintenance for Heating Ventilation Air Conditioner (HVAC) assets. My work has largely been in the intersection of ML and software engineering. You can read more about me here.
I am building an open source Machine Learning pipeline for structured data (python based) which can be used for any Data Science competitions as well for real life projects (at work). This project is work-in-progress
For newbie in ML, it will provide a project structure and open source APIs for end to end Data Science work (explore/visualize data, engineer features, train model, track experiments, select features, optimize hyperparameters, interpret models and make predictions). The code will follow the software development best practices (designed/developed following standard design patterns, well documented, unit tested, CI/CD enabled). As a result, when they are competing on any ML competition, they are learning about real software development too.
This project can be readily used for real work projects as well (since it follows standard software development practices). As a result, a Data Scientist can quickly change their gear from fun competition to real life work. Also, organizations who are getting started with ML can cut short their initial development cycle.
I am building it "in public". I am sharing my work in competition platforms like Kaggle and collecting feedback from the Data Scientists to improve further.
The sponsorship will help me to dedicate more time in developing this project, include for features which data scientists can use directly for fun or work.
Kaggle Pipeline for tabular data competitions
Language: Jupyter Notebook - Stars: 204