Open to opportunities

Pratik Dongre

@pratikdongre

Message

Data Scientist with expertise in ML and analytics solutions.

India

Message

What I'm looking for

I am looking for a role that fosters innovation, collaboration, and growth opportunities in data science.

I am a Data Scientist with 3 years of experience in delivering impactful machine learning and analytics solutions across healthcare and research domains. My proven track record includes developing NLP pipelines, generative AI models, and predictive algorithms that significantly improved data accuracy and reduced manual processes. I excel at transforming complex datasets into actionable insights through scalable data pipelines, utilizing cloud platforms like GCP and AWS, and employing BI tools such as Looker and BigQuery.

In my current role as a Business Operations Associate at CertifyOS, I have successfully assessed business performance by designing KPIs and evaluating key metrics, which informed risk mitigation strategies and optimized operational decision-making. I built scalable analytical frameworks and data models that automated 80% of reporting workflows, reducing manual reporting time by 60%. My collaborative approach with stakeholders has driven data-driven decision-making and operational efficiency across multiple product lines.

Previously, I worked as an Analyst where I leveraged generative AI models to enhance processes, achieving a 10% improvement in NPI attribution. My experience also includes backend development and database management at Deloitte USI, where I optimized database performance and contributed to new feature development. I am passionate about utilizing my skills to drive innovation and efficiency in data-driven environments.

Experience

Work history, roles, and key accomplishments

Current

Business Operations Associate

Current

CertifyOS

Apr 2024 - Present (1 year 3 months)

Assessed business performance by designing KPIs and evaluating key metrics, delivering actionable insights for risk mitigation and optimized operational decision-making. Built scalable analytical frameworks and data models using BigQuery and Looker, automating 80% of reporting workflows and reducing manual reporting time by 60%.

KPI Design Business Performance Analysis BigQuery Looker Data Modeling Data Pipelines Risk Mitigation

Analyst

CertifyOS

Feb 2023 - Mar 2024 (1 year 1 month)

Leveraged generative AI models to enhance name matching, improving NPI attribution by 10%, and led NLP-based model implementation for data cleaning and extraction, increasing data accuracy by 50%. Developed data products for healthcare provider categories, reducing manual efforts by 98% in credentialing and monitoring processes, and established robust data pipelines for aggregating raw healthcare

Generative AI NLP Data Cleaning Data Extraction Data Products Health Data Data Pipelines Data Analysis Reporting

Analyst

Deloitte USI

Jun 2022 - Jan 2023 (7 months)

Collaborated in backend development and database management for a healthcare client, utilizing .NET C# and SQL. Completed 6 new features and optimized database performance, increasing query efficiency by 5%.

Backend Development Database Management C# .NET SQL Database Optimization

Research Assistant

National Tsing Hua University

Jun 2021 - Apr 2022 (10 months)

Investigated astrophysical simulation datasets related to AGNs, supervised by Dr. Hsiang-Yi Karen Yang, contributing to the scientific understanding of turbulence effects on AGNs. Leveraged Python software 'YT Project' to analyze hydrodynamic simulations, quantified parameters like velocity structure function, and authored a research report.

Astrophysical Simulation Data Analysis Python YT Project Reporting

Research Intern

University of Hawaii

Jun 2021 - Dec 2021 (6 months)

Developed a deep neural network model using PyTorch, trained with n-body cosmological simulation snapshots, to estimate simulation initial conditions, drastically reducing computational time. Conceptualized an algorithm to calculate reverse displacement fields of n-body simulations, optimizing computational resources by 40% and reducing simulation runtime by 50 hours per dataset.