Arpit Agrawal

About Arpit Agrawal

I lead initiatives across three core competency areas : AI/ML, Power Platform apps & automation, and data visualization & insights — applied to solving operational problems and improving execution speed in enterprise environments.Key Projects : 🧠 AI / ML• Layout-Agnostic Table Extraction (Decimal Accuracy) – Built a multi-technology solution using pdfplumber, Camelot, Power Automate, and PyTesseract hOCR to extract tables with decimal precision from digital and scanned documents, enabling validation-ready outputs and auto-generated Word reports, while shifting consultant effort from report building to analysis and advisory work.• GenAI-Powered PDF Extraction & File Renaming – Used PyTesseract, spaCy, and GenAI to extract key fields from PDFs and rename files—saving hours/year.• Automated Data Labeling Across Domains – Applied prompt engineering with GenAI to automate labeling and improve consistency.• OCR & Serial Number Verification System – OCR-based solution for serial number verification supporting warranty and operations workflows. Power Platform Apps & Automation• Excel → Bespoke Report Automation – Automated reports from Excel data, including images, using PowerApps, Power Automate, Office Scripts, VBA, and PCF—saving 100+ hours/year.• Legacy Access → Power Platform Migration – Transitioned MS Access system to SharePoint + Power Apps + Power Automate for real-time, cloud-based data entry. Data Visualization & Insights (Power BI)• Operational & Project Intelligence Dashboards – Dashboards combining 16+ data sources for consolidated planning and analysis.• Power Query Data Preparation – Applied transformation steps to bring raw data into dashboard-ready format, reducing prep time from days to hours.• Advanced DAX for Non-Native Capabilities – Implemented DAX patterns enabling analytical and reporting behaviors not natively supported in Power BI.

Experience

1. Manager Data Science

   Marsh

   Sep 2024 — Present · Gurugram, IN

   Layout-Agnostic Table Extraction (Decimal Accuracy) – Built a multi-technology solution extracting precise tables and shifting consultant effort from report building to analysis- Developed a GenAI-powered automation pipeline using PyTesseract, spaCy, and GenAI to extract entities from PDF documents and rename files, reducing manual effort by hours annually- Automated generation of bespoke reports from Excel files including images, using Power Automate, Office Scripts, Microsoft VBA, and PowerApps Component Framework, saving 100+ hours annually- Leveraged prompt engineering and GenAI models to automate data labeling tasks across multiple domains, increasing consistency and annotation speed- Power BI Dashboards & Data Modeling – Built dashboards integrating 16+ sources, using Power Query and advanced DAX to enable non-native analytical capabilities and faster preparation- Designed advanced dashboards for transforming unstructured data into structured insights with multi-level aggregation, storyboarding, and app-like navigation.

Education

• Maria Assumpta Convent School

  Senior School Certificate Examination

  2010

• Liverpool John Moores University

  Master's degree, Data Science

• Bipin Tripathi Kumaon Institute Of Technology

  Bachelor of Technology - BTech, Mechanical Engineering

  2010 — 2014

• Maria Assumpta Convent School

  Secondary School Examination

  2008

• International Institute of Information Technology Bangalore

  Executive Post Graduate Programme, Data Science