Sneegdharoop Bhattacharyya

• Lennar Homes.

Worked for Lennar Housing, which is an esteemed construction company in the US that needed to build a robust MLOps platform for their organization's data science and ML use cases. The platform needs to be scalable and capable of improving the overall model lifecycle, supporting a variety of ML applications. Helped in development of CI/CD workflow with GitHub Actions, using AWS resources like Sagemaker, Lambda, and Glue, while setting up the infrastructure using Terraform.

• DRL Analytics Automation.

Worked for DRL as a consultant in building, deploying, and containerizing several application archetypes with Jenkins CI/CD automation pipelines. Responsible to understand each process flow and business requirements to deliver appropriate DevOps, DevSecOps, and DataOps solutions for different types of applications like Python-based, Java-based, NodeJS-based, etc. Use of different source control tools like Git and Bitbucket for appropriate archetypes to provide an environment-based branching strategy. Use of AWS Secrets, SonarQube, Dependency Track, Cosign, and Qualys to provide different kinds of pipeline securities related to code level, package level, artifact level, and container level vulnerabilities. Pushing the secured image into appropriate, required container repositories like AWS ECR or JFrog. Responsible for both server-level and serverless deployment strategies in AWS. Application monitoring for server-level deployments, used Dynatrace APM, and used REST APIs to get JSON data regarding the security metrics to visualize them in PowerBI.

• Internal R&Ds and POCs (PwC)

Worked for a few MLOps/DevOps use cases and POCs for PwC for different use case requirements, including but not limited to Azure DevOps, Azure ML, DVC, Terraform, Docker, Kubernetes, and AWS Lambda.

• Intro-Act Automated Prediction

Worked for Intro-Act which is a financial technology company on US stock market functionalities. Has worked on developing python-based machine learning scripts for data preparation, pre-process, scoring & predictions to give financial prediction output using Keras, H2O.ai, XGBoost and Scikit-Learn. Automation pipeline was created using azure data factory, data and scripts managed, extracted, processed, versioned and archived using azure datalake gen2, azure databricks.

• Littler Mendelson P.C.

Worked for Littler Mendelson P.C. which is an esteemed law firm which needed to build a solution that can predict their lateral hiring success. Has worked on the solution entails creation of an Un-Supervised Machine Learning solution to label and categorize their employees into three distinct performance categories. Performed data preparation through python scripts and developed different Un-Supervised Machine Learning models using Scikit-Learn to label and predict the laterally hired employee performance categories. Output clusters and labels are visualized using power BI dashboard.

• Internal R&Ds & POCs (Datacore)

Worked for a lot of data science use cases, POCs & research papers for Datacore through several years for different internal analytical product requirements including and not limited to data analysis preparation cleaning and preprocessing, time series forecasting, deep learning, statistical inferencing, synthetic data generation, various transfer learning models, NLP, data visualization, ML API creation for backend etc.

• DCS AI AutoML (In-House Product)

The key objective was to build a web-based saas platform which will bring the power of Automated Machine Learning and Analytics under its hood to quickly analyze the data and generate best possible solution for anyone, without the hassle of knowing complex algorithms, empowering business as well as individuals.