Rajat

Leading and mentoring a team of MLEs and Data Scientists to complete various Data Science Projects.  Leveraging machine learning for solving retail problems like – Customer Segmentation, Forecasting Demand, Pricing/ Promotion Analytics. Introducing the best practices for implementing Data Science techniques, Version Control, Data Science engineering, automation, etc.

• Communicating with the business and leadership to understand the exact business problems to be solved using data
• Introducing MLOps best practices for model deployment, model monitoring, alerts, etc.
• Automating the infrastructure through AWS cloud formation / Terraform and serverless deployments.

Select Environment & Tools Used: Python, Spark, Scala, Azkaban/ Airflow, SQL, Tableau, AWS (Sagemaker, EMR, EC2, EKS), Docker, Kubernetes, Terraform, CloudFormation, Serverless, MLOps

Created Large Volume Data Pipelines using Spark for several use cases. Used a mix of tools like Spark, Scala, Azkaban, etc. to finish and automate the flows. Wrote ETL pipelines for signal processing. Performed feature engineering and made the features available to various teams for campaign optimization purposes. Developed and deployed the gender prediction model to improve the known gender by 25%.

• Used Airflow / Azkaban to orchestrate the data flows to submit the spark jobs to EMR / Glue.
• Optimized spark jobs to reduce resource utilization and improve efficiency
• Developed and productized a CTR prediction model using Python to forecast the CTRs of different properties on the web and app.
• Tableau Visualizations: Created dynamic dashboards to depict actionable insights from Analytics

Environment & Tools Used: Spark, Scala, Python, Azkaban/Airflow, SQL, Data Studio, CI/CD (Jenkins), Docker, EMR, Glue, Tableau, GCP, S3, Redshift, Big Query.

Worked with customer and site segmentation implementing various end-to-end solutions for clustering problems.  He applied Machine Learning algorithms to predict the likelihood of customers responding to a campaign and to launch nationwide customer loyalty campaigns. Built predictive models to reduce churn and upsell premium products. He regularly presented Data Science capabilities, research, and analysis at the CEO level.

• Pricing and Promotion Analytics: Used time series techniques to find out the impact of price changes for different sets of customers
• Text Analytics: Sentiment Analysis and Topic Modelling to identify the drivers of various products using the customer reviews
• Implemented Market Basket Analysis, Network Analysis, SKU, and  Assortment for a retail client
• Applied knowledge of ML and AI techniques for Sales Datasets
• Used topic modeling (NLP) to automatically classify reviews into various topics and then used sentiment analytics on top of it to ascertain the public’s opinion for that SKU and topic.
• Automated several manual workflows saving considerable time for supply chain and marketing analysts.

Environment & Tools: SQL / Database, R, Python, Tableau, Advanced Excel, PowerPoint, Java, Unix, NLP, XGBoost, Random Forest, Prototyping, Flask, AI, MySQL, Data Lakes, PowerBI, Redshift, S3, EC2, PostgreSQL

• Corporate Trainer, Data Science at Educonverge (2018)
• Intern at Accenture AI (2015)
• Research Assistant at National Food Security Act 2013 (2014)
• Founder at Dunomics (2011-2012