Apurva Tawde

About Apurva Tawde

I am a Machine Learning Engineer II at Cohere Health, a leading company that leverages AI…

Experience

1. Machine Learning Engineer Ii

   Cohere Health

   Jan 2020 — Present

   The implementation of end-to-end functional & operational monitoring for ML models led to enhanced model performance, improved model reliability, and quicker issue detection and resolution. •Led Proof of Concept (POC) for ML Model Observability platform, enhancing model reliability, performance, and troubleshooting across a diverse range of 10 vendors. Conducted impactful experiments to validate efficacy. •Successfully collaborated with cross-functional teams to set up the necessary data pipelines, ensuring seamless integration with the Arize platform, and enabling real-time monitoring capabilities. •Executed note-based info extraction experiments using novel LLMs via prompt engineering MLOps: •Crafted and implemented strategic MLOps roadmap, driving agile ML model development and deployment enterprise-wide. •Established streamlined ML-Ops guidelines, yielding smooth releases, error reduction, and superior control. •Generated actionable cost reports, driving data-driven decisions for resource allocation and budgeting. •Specialized in MLOps, elevating Cohere\'s operational efficiency and model performance. Hiring and Leadership: •Conducted interviews for experienced ML Engineers and Senior MLOps Engineers, enhancing team expertise and outcomes. •Empowered DS/ML teams by enabling them to utilize the Platform effectively through well-organized and tailored training sessions. •Built ground-up data pipeline for ingesting health insurance claims and member data from multiple providers using complex SQL scripts in AWS Athena •Implemented techniques like forwarding selection, backward elimination, and stepwise approach for the selection of the most significant independent variables Implemented Cost Ranking Model using regression models to optimize service operations using historic claims data and product data to predict ROI •Employed NLP extraction for patient-level features, like clinical data (smoking, BMI) via classifiers from clinical notes.

Skills

• Software Development
• Sql
• Android Development
• Databases
• Tibco Bw 6
• Eclipse
• Linux
• Microsoft Word
• Photoshop
• Powerpoint
• Php
• Teamwork
• Windows
• Javascript
• Core Java
• Social Media
• Winscp
• Xml
• Android
• Html
• Public Speaking
• Tibco Ems
• Java
• C
• C++
• Mysql
• Tibco
• Tibco Businessworks
• Microsoft Excel
• Microsoft Office