Hi, I'm Heali Mehta

Data Scientist & Machine Learning Engineer

Master’s student in Computer Science at Texas State University with a focus on Data Science. Driven by a passion for transforming data into actionable insights, I specialize in building intelligent systems that solve real-world challenges. My experience spans diverse projects in computer vision, natural language processing, and predictive analytics, fueling my growth as an aspiring AI/ML engineer.

About Me

Driven by curiosity and powered by data

Hello! I'm Heali, a Master's student in Computer Science at Texas State University with a concentration in Data Science. My journey has been shaped by both academic and professional experiences, through which I discovered my passion for leveraging data to solve complex problems.

Recently, I worked as a Data Engineer at Shreem Solarium and a Business Intelligence Analyst at Saptarishi Furniture, where I developed machine learning models, automated workflows, and supported data-driven business decisions. I’ve been deepening my expertise in computer vision, natural language processing, and deep learning using PyTorch, TensorFlow, and cloud technologies.

Outside of tech, I enjoy exploring emerging technologies, engaging in research projects, and embracing lifelong learning. Life, like data, becomes more fascinating when you’re curious and open to discovering new patterns.

93%
CNN Model Accuracy
85%
Recommendation Precision
5000+
Data Points Analyzed
3rd
Place Datathon 2025

Professional Experience

Building expertise through hands-on projects and research

Texas State University
Graduate Research Assistant
San Marcos, Texas, United States · On-site
Jan 2024 – May 2025
Collaborating on advanced data mining research projects focusing on text and semantic analysis.
  • Developed a web app using Python, Flask, PHP, and HTML/CSS for advanced lyric-based document retrieval
  • Achieved 85% retrieval accuracy through TF-IDF, context-aware search logic, and ranking techniques
  • Processed and analyzed data across 5,000 songs with optimized text preprocessing methods
Shreem Solarium
Data Engineer
Vadodara, Gujarat, India · On-site
Sep 2022 – May 2023
Engineered machine learning solutions for solar panel maintenance prediction and optimization.
  • Built Random Forest model for solar panel maintenance prediction achieving over 85% accuracy
  • Utilized Docker and Git for containerization and version control of model deployments
  • Developed Flask REST API for real-time predictions integrated with Laravel web applications
Saptarishi Furniture
Business Intelligence Analyst
Vadodara, Gujarat, India · On-site
Jan 2022 – Aug 2022
Developed intelligent recommendation systems and automated analytics pipelines for e-commerce optimization.
  • Developed TF-IDF-based product recommendation system using cosine similarity
  • Improved cross-sell precision by 80% and enhanced decision-making for marketing teams
  • Built SQL and Python pipelines to automate workflows into analytical dashboards
YokeMate
Machine Learning Intern
Vadodara, Gujarat, India · Remote
Jan 2021 – Nov 2021
Created data-driven modules for food delivery system optimization and user experience enhancement.
  • Developed data-driven modules including login systems, delivery logic, and chatbot integration
  • Worked in Agile sprints, increasing team velocity by 15% through effective backlog grooming
  • Contributed to building scalable food delivery platform with intelligent features

Featured Projects

Solving real-world problems with data science and machine learning

Skin Cancer Detection System

Python, TensorFlow, CNN, Computer Vision

Developed and trained a Convolutional Neural Network model for classifying skin lesions with exceptional accuracy metrics. The system demonstrates the potential of AI in medical diagnosis and early disease detection.

93.59% Accuracy 92.14% Sensitivity 95.22% Specificity

Movie Recommendation System

Python, TF-IDF, Cosine Similarity, NLP

Developed a content-based movie recommendation engine utilizing TF-IDF vectorization and cosine similarity, effectively analyzing a dataset of over 5,000 movie plots to provide personalized recommendations.

85% Precision 5000+ Movies Content-Based

Generative AI Model

Python, OpenAI API, Fine-tuning

Gained hands-on experience with OpenAI APIs and fine-tuned model performance for specific text generation tasks while achieving optimized response times for real-world applications.

<1.2s Response OpenAI Integration Fine-tuned

Lyric-Based Document Retrieval

Python, Flask, PHP, TF-IDF, Search Algorithms

Advanced web application for keyword-in-context search across large music datasets using sophisticated text processing, ranking algorithms, and context-aware search logic.

5000+ Songs 85% Accuracy Context-Aware

Skills

Technologies and tools I’ve worked with

ERP & Business Systems

Baan ERP SAP Hybrid Cloud Infrastructure (AWS) Application Maintenance System Implementation Requirement Gathering Change Management User Acceptance Testing (UAT)

Programming & Scripting / Automation

Python C++ Bash Shell Scripting Linux System Programming SQL (complex queries, data integration, reporting) Git Automating IT Operations Process Workflow Optimization

Data Analysis & Visualization / Reporting

SQL Redshift Data Lakes Pandas NumPy Excel Matplotlib Seaborn Tableau TensorBoard ETL Pipelines BI Dashboards Service Level Agreements (SLAs) Data Migration

Machine Learning & Deep Learning

Scikit-learn PyTorch TensorFlow Keras JAX (basic) OpenCV Random Forest Logistic Regression Time Series Clustering TF-IDF A/B Testing Computer Vision (CNNs) NLP (LLM fine-tuning) Recommender Systems

ML Engineering & MLOps

Docker Airflow MLflow FastAPI Flask REST APIs CI/CD AWS (Glue, S3, EC2, Lambda) Kubernetes (basic)

Cloud & IT Infrastructure

AWS Certified in Machine Learning – Specialty Hybrid Cloud Monitoring System Backup & Retention Proactive System Monitoring Disaster Recovery Planning and Testing

Systems & Architecture

GPU Acceleration CUDA Model Quantization Pruning Memory Profiling Compute Profiling

Tools & Frameworks

PyTorch Lightning XLA Triton

System Administration

Install, Configure & Administer Application Systems Technical Specifications System Enhancements (Software/Hardware Updates) Troubleshooting & Vendor Communication

Security & Compliance

Data Integrity Data Security (Cloud & Enterprise Environments)

Project Management & Collaboration

Standard Operating Procedures (SOPs) Budgets & Scheduling Performance Metrics Feedback Channels Cross-team Communication Customer Communication

Achievements

Recognition for innovation and excellence in data science

3rd Place – Datathon 2025

Healthcare Distribution Analysis in Austin

Led a team of 4 in TXST Datathon 2025, where we analyzed the spatial distribution of hospitals in Austin relative to population density. We developed an interactive dashboard and performed geospatial analysis using datasets from the Texas State Data Repository, delivering actionable insights to enhance healthcare accessibility. Our project secured 3rd place in the graduate division.

Contact

Let’s connect and build something meaningful

Feel free to reach out for collaboration, freelance projects, or just a friendly chat about all things data and machine learning. I’m always open to exciting opportunities and meaningful conversations!