Product Data Scientist
July 2024 — July 2025At GenerativeProducts
Hey! My name is Prajwal and I was born in Bangalore, India.
🎓 Recently, I graduated from Master of Science in Business Analytics & Artificial Intelligence at The University of Texas At Dallas after 2 long years.
🏢 I have previously interned at GenerativeProduct.io during 2024 summer as a Product Data Scientist. This was my first time working with LLMs, RAGs and Agentic AI.
🎮 Beyond the Code – Passionate about Current Affairs & News, Tech Trends, and Football.
I am a Data & AI Engineer with a strong foundation in data science, business intelligence, and machine learning. My work involves building scalable data models, designing insightful dashboards, and developing ML solutions — all while aligning closely with business objectives. I approach every data challenge with ownership and a commitment to practical, effective outcomes.
Contact me at prajwalkp.work@gmail.com
I'm currently diving deeper into generative AI and large language models, exploring how they can transform business intelligence and analytics.
Exploring Microsoft Synapse Analytics and Azure services to build scalable data solutions, optimize analytics workflows, and enhance cloud-based data processing.
Languages
Python R SQL SAS Bash/Shell ScriptingTools
Data Visualisation - Tableau, Looker and PowerBI Databricks Snowflake Container Management - Docker and Kubernetes Version Control - Git, GitLab and Azure Repos Jira Apache Kafka, Airflow, Spark HadoopCloud
Microsoft Azure - Synapse Analytics, DevOps/Pipelines AWS – EC2, S3, QuickSightLibraries
Numpy Pandas Matplotlib Plotly Seaborn Flask and DjangoBuilt an automated ETL pipeline using Apache Airflow to extract book data directly from the Amazon Bookstore, transform it for consistency, and load it into a PostgreSQL database. This enabled structured analysis of titles, pricing, ratings, and availability for downstream insights.
CometVerse is a voice-enabled AI chatbot built for UT Dallas, designed to handle student queries using LLMs and NLP. It supports conversation history, voice input, and PDF export, offering a smarter, more interactive alternative to traditional university chatbots.
An AI-powered tool designed to help candidates prepare for Product Management interviews by providing automated real-time feedback. It leverages Speech-to-Text, Retrieval-Augmented Generation (RAG), and LLMs to enhance interview readiness and improve responses.
Developed a predictive model using ML techniques (Linear/Polynomial Regression, SVR, Decision Tree, Random Forest, etc.) to forecast ride demand. The project supports smarter driver allocation, pricing strategies, and overall efficiency in ride-hailing services.
To build this, we will make usage of a popular distributed streaming platform Apache Kafka, to produce, consume and stream our data into BigQuery (data warehouse) so we can do some analysis using SQL and connect it to PowerBI where we are going to visualize on a dashboard.
The goal was to understand player injuries based on specific variables related to plays & game conditions. Apart from basic ML models, traditional survival analysis models were also explored - Cox Proportional Hazard model. Feature Importance using Mean Decrease in Impurity (MDI) was obtained.
Presented important insights and built a model for the prediction of real estate prices in the city of Bengaluru. Got an accuracy score of 0.68 for the model used. Comparison of other models - Lasso and Decision Trees was also explored.