Nefeli Zafeiri
  • Home
  • Capstone Project
  • GitHub
  • Hugging Face
  • LinkedIn
  • Email
Deloitte Capstone, Jan 2026 to May 2026

AI-Driven Wildfire Risk Monitoring

Translating satellite data and machine learning into real-time early-warning insights for wildfire response teams.

Python Machine Learning Geospatial Data Data Visualization Deloitte Partnership

Overview

As part of the University of Miami MSBA Deloitte DIRPA cohort, our team partnered with Deloitte to design an AI-driven solution for wildfire risk monitoring and early detection. Wildfires cause billions in damages annually and early-warning capabilities can directly save lives. This project bridges advanced analytics with operational decision-making for emergency response stakeholders.

My Role

I served as a data analytics and strategy contributor, responsible for translating analytical outputs into actionable business insights. My focus spanned data pipeline exploration, stakeholder communication, and framing model outputs as decision-ready intelligence.

  • Analyzed environmental and satellite datasets to identify fire-risk indicators
  • Developed data visualizations to communicate risk patterns to non-technical stakeholders
  • Contributed to the project strategy and client-facing reporting framework
  • Collaborated across a cross-functional team with Deloitte advisors

Project Scope

Client Deloitte
Program DIRPA, Deloitte Institute for Research and Practice in Analytics
Duration January 2026 to May 2026
Domain Environmental Risk, AI / ML, Public Safety
Live Demo wildfirewatch-app.streamlit.app
Repository github.com/nefelizafeiri/wildfirewatch-app

Approach

01

Data Collection and Exploration

Gathered and cleaned geospatial, weather, and historical wildfire datasets to build a reliable analytical foundation.

02

Risk Modeling

Applied machine learning techniques to identify environmental risk factors and predict high-probability ignition zones.

03

Visualization and Insight Design

Designed interactive risk maps and dashboards enabling stakeholders to interpret outputs without specialized data science backgrounds.

04

Strategic Recommendations

Translated model findings into operational recommendations, aligning analytics outputs with real-world emergency response workflows.

Key Takeaways

🔥

Real-World Impact

Directly applicable to emergency management decision-making at a scale where timing matters most.

🤝

Industry Collaboration

Worked alongside Deloitte professionals, gaining hands-on exposure to enterprise-level consulting workflows.

📊

Insight Communication

Refined the ability to present complex analytics in a way that drives action across diverse audiences.

Back to Portfolio

© 2026 Nefeli Zafeiri, nefelizafeiri@gmail.com, LinkedIn