I am a Data Science student driven by a single mission: Finding the "blind spots" in our systems and building the technical bridges to fix them. Whether it’s healthcare access, financial equity, or career transitions, I use data to ensure underserved groups aren't left behind.

• 1st Place Team Winner – HeatMap Hackathon (03/2026): Ranked #1 in the state for identifying geographic disparities in U.S. burn care access. Managed the technical GitHub repository and ensured project integrity through data-driven storytelling.
• 1st Place Winner – Minnesota AGENT.AI Competition (10/2025). Ranked #1 in the state for predictive accuracy and technical logic in retail banking fraud detection.
• Dean’s List & Scholarship Recipient – Recognized for consistent academic excellence.

• AgentDS-Retail-Banking-Fraud: The real-world data and domain knowledge & logic and Python pipeline behind the 1st Place MN AGENT.AI win.
• HeatMap-Burn-Care-Access): Collaborative Hackathon project. I developed interactive mapping and data synthesis to identify gaps in healthcare services.
• Developer-Skills-Analysis: IBM Capstone. Engineered a pipeline for 70k+ records to bridge the gap between education and market demand.
• ai-form-analysis: 3-Month Technical Roadmap: Built OpenCV + MediaPipe pipelines for pose detection and frame sampling from MP4 data. Extracted joint angles and stored annotated frames with summary data.Developed biomechanical analysis across 5 throw phases — compared joint angles against elite athlete profiles (flagging >15° deviations), estimated hip/wrist rotational velocity, and integrated Gemma vision models to generate structured coaching reports.,Shipped a Streamlit web app combining the full pipeline into a single interface. Users upload video and receive pose overlays, velocity metrics, and AI coaching feedback — running entirely locally via Ollama (no API costs, no internet required).

While the broader team focused on the Composite Vulnerability Index (CVI) presentation, this work provides the underlying financial and geospatial engine — quantifying state-level avoidable hospitalization costs from burn under-referral, scoring 498+ trauma centers as telemedicine hub candidates, and mapping population-weighted distance burdens using Census and NIRD 2023 data.

• Economic Insight: Modeled avoidable hospitalization costs from burn under-referral at the state level using Murray et al. (2019) excess LOS data ($3,500/day) and Huang et al. (2021) 66% under-referral rate, producing state-by-state cost burden estimates from the NIRD 2023 database.
• Innovation: Engineered a "Tele-Burn Hub" opportunity scoring system ranking non-burn trauma centers by referral gap potential, factoring in trauma level, bed capacity, and burn care absence — mapped geographically using Folium.
• Impact: Identified high-priority telemedicine spoke candidates (opportunity score ≥ 5) across states with the greatest access-to-quality gaps, with projected reductions in excess hospitalization days and avoidable costs.
• Tech: Python, Folium Geospatial Mapping, ROI Modeling

• [Resource-Hub]: "Humanity First" system .

I maintain a structured "Just-in-Time" learning library to bridge the gap between academic theory and industry-grade execution.

Focus: Mastering autonomous systems and prompt engineering.

• ai-agents-for-beginners - Microsoft's 12-lesson curriculum on agent architecture.
• GenAI_Agents - Advanced implementations for intelligent, interactive AI.
• Hands-On-LLMs - Practical application of Large Language Models.

Focus: Scaling pipelines and professional visualization.

• IBM-Data-Analyst-Capstone - Analyzing tech trends using Python and SQL.
• Meta-Database-Engineer - End-to-end database design and optimization.
• Google-Advanced-Analytics - HR turnover prediction using advanced predictive modeling.

Focus: Understanding the "why" behind the models.

• nn-zero-to-hero - Andrej Karpathy's neural network fundamentals.
• annotated_DL_papers - 60+ implementations of deep learning papers.
• GPU-Puzzles - Learning CUDA and low-level optimization.

📫 Connect with me: LinkedIn | GitHub