Vikas Singh Kaviya

About Vikas Singh Kaviya

Data scientist with 5+ years of experience in statistical analysis, machine learning, natural language processing and data manipulation. Proficient in Python for insights extraction, model building, and deployment. Skilled communicator driving data-driven decisions, with a keen interest in integrating trending AI developments.

Experience

1. Senior Data Scientist

   TartanHQ

   Apr 2023 — Present

   Tartan is a fintech platform offering consent-driven data exchange solutions, facilitating real-time verification of work and income- Led the development of data science and analytics initiatives in the fintech sector- Fine-tuned Layoutlmv2 for payslip data extraction, achieving 72% accuracy, surpassing the baseline model- Utilized hyperparameter tuning to train LLM for document classification, achieving 84% accuracy- Built a blur document detection model with AutoML and OCR tools- Customized labelling tool for training data preparation, Developed AWS deployment-ready inference endpoints- Implemented prompt engineering techniques to optimize GPTs\' performance across diverse tasks- Utilized analytics for marketing campaign hypothesis testing and product line analysis, driving a 20% increase in leads- Integrated Apache Superset, an open-source data visualization tool, developed various visualisation dashboards and automated periodic MIS email alerts- Developed Flask applications for data extraction with Selenium and crawling libraries, and Python script for record matching using record-linkage modules

Education

• The LNM Institute of Information Technology

  Bachelor of Technology (B.Tech.), Computer Science and engineering

  2014 — 2018

• Indira Gandhi National Open University

  Post Graduation Diploma in Applied Statistics, Statistics

  2021 — 2022