All Projects
Dive: Agentic AI · Weather Intelligence
Vesper AI
A mobile sky-forecasting app to recommend and visualize sunrise/sunset viewing spots against user preferences.
Jul 2026
- Predicts sunset/sunrise quality using cloud cover, visibility, humidity, wind, and atmospheric conditions
- Recommends Top 5 viewing locations based on forecast and geographic conditions
- Supports persistent or session-based preferences and condition-triggered notifications
Dive: Agentic AI · Location Intelligence
Rel-AI
An agentic system to help recently relocated people rebuild their fitness, workspace, and outdoor routines.
Aug 2026
- Discovers and compares nearby gyms, Pilates studios, classes, and running routes
- Combines location, pricing, reviews, weather, and user preferences for personalized recommendations
- Maintains preference memory and supports fitness scheduling/calendar integration
Dive: Healthcare AI · Machine Learning
Surgical Copilot
Developed a clinical decision-support application that predicts surgical risk from preoperative patient data using machine-learning models.
Jul 2026
- Trained and evaluated multiple gradient-boosting models for preoperative risk prediction
- Added SHAP-based feature attribution to explain individual predictions
- Built REST APIs serving serialized ML models with real-time frontend prediction visualization
Dive: Multimodal AI · Mental Health
Depression Detection
Developed a multimodal machine-learning system that combines textual and visual signals to detect patterns associated with depression.
Apr 2024
- Combines textual and visual inputs through a multimodal deep learning pipeline for depression classification
- Uses BERT, TF-IDF, NLP preprocessing, and image-processing techniques to extract predictive features
- Achieved 82% classification accuracy and deployed model inference through a Flask-based web application
Dive: Operations Research
Airline Optimization
Built Python-based optimization models to evaluate flight allocation scenarios across 30 airline routes.
Apr 2026
- Optimized flight frequency and capacity allocation across a network of 30 airline routes
- Built integer programming models using Gurobi and Hexaly to evaluate route-level allocation scenarios
- Increased projected network profit by $43M (7.47%) while achieving approximately 85% capacity utilization
Dive: Financial Analytics
Pizzeria Analysis
Queried, cleaned, and normalized operational and financial data, then built decision-support models.
Oct 2025
- Cleaned, normalized, and analyzed operational and financial data to evaluate business performance
- Built NPV, IRR, and payback models to assess the financial viability of potential expansion scenarios
- Combined revenue, cost, cash flow, and investment projections to support data-driven expansion decisions
Dive: Big Data
Hospital Operations Dashboard
Cleaned, modeled, and visualized healthcare data to uncover trends from 9M+ clinical records.
Jun 2026
- Processed and modeled 9M+ clinical records to analyze hospital utilization and operational performance
- Built interactive Power BI dashboards tracking admissions, bed occupancy, length of stay, readmissions, and mortality
- Used SQL, Power Query, and DAX to transform clinical data into operational trends and decision-support metrics
Dive: Data Visualization
Climate Policies Dashboard
Built an interactive dashboard to analyze U.S. climate policies and compare state-level initiatives from 2004-2024.
Apr 2025
- Analyzed U.S. environmental policy activity across states and regions spanning 2004–2024
- Built an interactive Tableau dashboard for exploring policy distribution, adoption trends, and geographic differences
- Enabled time-series and regional comparisons to identify patterns in environmental policy activity across the United States
Dive: Business Analytics
Capstone Consulting
Built a recommendation engine to provide vendor recommendations with the adoption strategy.
Sep 2025
- Evaluated eight MBSE platforms using a weighted decision framework covering usability, interoperability, cost, and simulation capabilities
- Developed vendor recommendations supported by comparative scoring, cost-benefit analysis, and stakeholder requirements
- Created a phased change-management roadmap covering implementation priorities, governance, training, ownership, and adoption
Dive: Operations Research
Housing Hub Location Model
Built a p-median model to identify optimal hub locations under capacity constraints.
Feb 2026
- Built a capacitated p-median optimization model to determine optimal housing hub locations across candidate sites
- Minimized aggregate travel distance while accounting for hub capacity constraints and geographic demand distribution
- Used Python and Hexaly to solve and compare location-allocation scenarios for data-driven facility placement
Always Building..
Next project coming soon.
“Great work is never finished, only picked back up — one build, one idea, one late night at a time.”