Ayush Gupta

Projects

Medical Affairs

• Identified medical unmet needs and patient cohort for three disease areas based on business rules and generated patient physician mapping to be targeted by field reps.
• Used Kedro and pyspark to develop the data engineering pipelines.
• Managed the project in both development and support phases.

APLD Processed Claims

• Processed raw claims data in-house (on client environment) replacing processed data purchase from IQVIA.
• Offered flexibility of customization, speed, efficiency, accuracy and pursued cost savings of $3M annually.
• Processed claims data consumed by field sales team to identify patient activity, physician demographic details and treated patient claims.
• Used AWS Glue, Lambda, Athena, S3 and Pyspark to build the solution.
• Managed the engagement starting from requirement gathering to development and support.

AWS Architect

• AWS infrastructure setup and support role using services like Redshift, EMR, IAM, Lambda, RDS, EC2, S3 and Cloudwatch.
• Formulated an extensive cost optimization plan and successfully reduced the infrastructure cost by $800 per month.

JET Wave 2 Development

• JET is a data science model designed to enhance the experience of sales rep and healthcare professional's partnership.
• The model generates personalized suggestions for the sales rep to target the HCP based on various attributes such as specialty to make HCPs aware of the latest products that will eventually increase revenue.
• Responsible as a data engineer lead to communicate with different stakeholders to gather requirements, design and develop the pipelines for data engineering processes using Python, Pyspark and dataiku.

JET Operations

• Delivered the call and email suggestions generated using machine learning models based on the historical connects with the physicians and the resultant sales.
• Represented the team in the client meetings for regular deliverables, issues and enhancements.
• Created different features for the Health Care Professionals based on different attributes.
• Used Pyspark, Dataiku, AWS S3 and Redshift to meet the objective.

Omnichannel Development

• Developed a common platform 'Omnichannel' for marketing and sales data for pharmaceutical client to get the trends and insights on a single platform.
• Gathered requirements and business rules from respective data scientists and created triggers for sales reps.
• AWS Redshift, Hive, Dataiku and Python were used in the development.

Hadoop and Linux Server Administration (AWS)

• Worked as Hadoop Administrator in data warehousing project. Created and maintained more than 10 AWS EMR clusters with up to 120 nodes.
• Optimized AWS Infrastructure used in the project with Auto-scaling and HDFS, Yarn, Tez and Spark tuning significantly by almost 38%.

Dataiku Infrastructure Implementation and Support

• Created infrastructure platform for Data Science and Engineering projects to be developed on Dataiku.
• Docker containers have been deployed to offload the python jobs on dedicated machines and keep the environments uniform on each container.
• EMR clusters have been introduced to cater Pyspark jobs.
• Involved users and access control management for current users and new onboarding.

Talend and EMR Upgrade

• Worked on upgrading Talend ETL tool and AWS EMR, consisting of Hive & Spark Technologies.
• Solved various compatibility issues like snappy compression version conflict, enforced removal of redundancy in Tez framework.

Data Warehouse Design and Development

• Developed a Hadoop based data warehouse using the ETL tool Talend, Pyspark and Hive query language.
• Migrated the traditional “Oracle” based data warehouse for a US pharma major client.