Harsh Mehta
Founder & data scientist turning raw data into packed venues. Building Ensemble, the AI-agent booking engine. Business-intelligence analyst by trade, AI strategist by obsession. Music connoisseur, live audio engineer, cinephile, spiritually Sanatani Hindu and history buff. INTP hoarding datasets on a self-built NAS while tinkering with MCP and next-gen agents. If it sounds, stores, or scales, I'm in.
About Me
I am Harsh Mehta, a data scientist, founder, and builder focused on mastering AI through code, systems design, and disciplined iteration. I build intelligent automation, machine learning products, and data infrastructure that solve real business problems. My work at Ensemble AI, HP Tech Ventures, and Beats by Dre reflects a balance of technical depth and strategic execution.
Beyond data, I explore music visualizations, cinema, cultural patterns, and creative systems. I help solopreneurs and small teams scale faster through automation and sharp problem solving.
Education
University of Wisconsin-Madison
M.S. in Information (Data, ML, Cloud Focus)
Sept 2023 - May 2025
GPA: 3.95/4.0
University of Mumbai
Bachelor of Management Studies (Business Analytics Major)
Jun 2013 - Jul 2017
GPA: 3.5/4.0
Experience
As a Founder, Data Scientist, and Analyst, I apply deep expertise in data science, machine learning, and cloud technologies (AWS, Python, SQL) to uncover strategic insights and tackle complex business problems.
Ensemble AI
AI powered venue booking platform using agent based architecture (n8n, A2A)
HP Tech Ventures
Evaluated startups for investment by analyzing data with Python and developing Snowflake ETL framework
Beats by Dre
Led competitive analysis and gathered consumer insights using Python NLP to boost marketing effectiveness
Prayas Entertainment
Boosted operational efficiency and revenue through data analytics and predictive modeling
Indigo Events & Promotions
Directed data-driven media strategies that improved customer satisfaction and lead conversions
UW Transportation Services Campus Parking Efficiency Project
Analyzed parking transaction records to improve resource allocation and reduce search time
Projects
A selection of my recent projects in data science, machine learning, and cloud computing.
UW Transportation Services Campus Parking Efficiency Project
Analyzed 11M+ parking transaction records using Python, SQL, and Snowflake to uncover usage patterns and weather impacts.
MSBA Financial Group Cloud-Native Data Architecture Project
Designed an end to end AWS data pipeline (S3, Glue, Redshift).
Hard Drive Data Extraction Tool
Developed a user friendly tool (Python, JS, PostgreSQL) to extract hard drive data to JSON, standardizing output for analytics.
Research: Job Posting Analytics on Twitter
Co authoring research analyzing 100k+ Twitter job postings using data mining and NLP to uncover hiring trends for peer reviewed publication.
Creative Application: MIDI to CSV Data Tool
Created a Python tool to transform MIDI music files into analytical CSV datasets, capturing musical structures for analysis in a creative domain.
Graduate Studies: Learnings & Reflections
Overview of key concepts, skills, and insights gained during the M.S. in Information program at UW Madison.
Skills
My technical expertise and professional capabilities.
Blog
Thoughts and insights on data science, AI, and technology.
Gallery
Visual showcase of my projects and work.
Data Visualization Dashboard
Interactive dashboard created for UW Transportation Services showing parking usage patterns.
Ensemble AI Platform
The Ensemble AI venue booking platform interface showing the agent-based architecture in action.
Financial Analysis Model
Predictive model for financial forecasting developed for Prayas Entertainment.
NLP Sentiment Analysis
Visualization of sentiment analysis performed on customer reviews for Beats by Dre.
Cloud Architecture Diagram
AWS cloud architecture designed for the MSBA Financial Group project.
Data Mining Results
Results from data mining analysis showing customer segmentation patterns.
Startup Analysis Dashboard
Dashboard created for HP Tech Ventures to evaluate startup investment opportunities.
Machine Learning Model Performance
Performance metrics for bankruptcy prediction model with 99.19% accuracy.
Get in Touch
If you are hiring, investing, or collaborating, I invite you to explore and connect.