PDF Data Extraction: Working mainly on extracting tabular and unstructured data from different financial statements.
Extracting unstructured data from tables like Assets and Liabilities, Operations, Changes in Net assets, Investment Summary, and Notes to Investments sections from the financial statements.
Using Python scripting, Pretrained Deep Learning Model, Azure Form Recognizer (Azure Cognitive Services), and Generative AI to extract the sections based on the scenario.
Has hands-on experience on GPT models like 3.5, 4v, and 4o models and has completed multiple use cases with proper prompting techniques and achieved great accuracy.
Modified the PDF Extracted outputs into database consumable format for easy reconciliation of the financial values, which was very effective for one of the Reputed Clients, saving a huge amount of manual validation of PDFs.
Extensively processed over 1,500 PDFs and achieved nearly 95% accuracy in extracted data.
Developed more than 10 validation scripts to validate the extracts and compare between drafts.
Containerized the developed Python automation script using Docker and scheduled it to trigger at intervals using a cron job.
All code was reviewed, perfected, and pushed to production.
Data Science Intern
Codevector Labs
Banglore
03.2021 - 08.2021
Built machine learning models for classification and regression problems.
Created visualizations of data with Tableau, PowerBI, and R Shiny.
Analyzed customer behavior using predictive analytics techniques such as regression, clustering, and decision trees.
Education
Electrical and Electronics Engineering -
Cochin University of Science and Technology
Kochi
01.2021
High School Diploma -
Jawahar Navodaya Vidyalaya
Kasargod
05-2016
Skills
Python
SQL
Generative AI
Git
Docker
Azure Databricks
Microsoft Form AI
Power BI
Excel
Machine Learning
Accomplishments
Participated and reached Top 10 in EYDeate Hackathon, where more than 600 teams from 5 countries participated. We built an end-to-end data extraction to reconciliation prototype using AzureFormAi models and Python PDF extraction libraries.
College project won 2nd in Myron Zucker International Project Design Contest as part of IEEE IAS CMD Awards - All Directional Robot
Won Goal Corporate Football Tournament- Gold Medal, Played for EY Kochi