Leonardo Calle, Ph.D.

PROFESSIONAL SUMMARY

Technical leader with 15+ years building production ML systems for complex temporal and spatial data analysis. Proven track record scaling algorithms from research through deployment, with expertise in risk modeling, signal processing, and leading cross-functional teams. Recently founded two AI companies while maintaining deep technical expertise in physics-informed machine learning and earth system modeling.

BUSINESS IMPACT

Production Systems: Deployed ML algorithms processing continental-scale datasets in high-performance computing environments with rigorous validation protocols.
Cross-Functional Leadership: Led interdisciplinary teams across NASA, USFS, and private sector partnerships, managing complex stakeholder relationships spanning government, academic, and industry partners.
Algorithm Development: Built end-to-end ML pipelines from requirements specification through production deployment, specializing in risk prediction and temporal pattern analysis.
Regulatory Experience: Delivered systems meeting NASA validation standards and federal agency compliance requirements.

TECHNICAL EXPERTISE

Machine Learning & Production Systems

Production ML frameworks: PyTorch, TensorFlow, FastAPI, PostgreSQL
Advanced architectures: Transformers, ViTs, LSTMs, CNNs
MLOps: Docker, CI/CD, automated testing, monitoring
Cloud & HPC: AWS, distributed computing, high-performance computing

Risk Modeling & Analytics

Predictive modeling for wildfire simulation and real-time risk assessment
Continental-scale classification & forecasting systems
Time-series analysis, statistical pattern recognition, multi-modal sensor integration
Physics-informed neural networks integrated with traditional simulation approaches

Software Engineering

Production systems in Python; Git workflows; testing & documentation standards
Database management (PostgreSQL), large-scale data pipelines
System reliability: performance monitoring, failure analysis, validation

PROFESSIONAL EXPERIENCE

Founder & CTO, mewmewmew Inc.

January 2025 – Present

Production AI Platform: Architected and deployed FastAPI-based ML system with transformer models serving an active client base.
Algorithm Engineering: Built complete ML pipeline from data preprocessing to deployment, specializing in document analysis & NLP.
System Reliability: Implemented automated testing, monitoring, and deployment workflows ensuring consistent performance.
Technical Leadership: Led interdisciplinary team through full product lifecycle from research to production.

Founder & CEO, CalleEcology Inc.

January 2025 – Present

Product Development: Created AI-integrated curriculum teaching modern deep learning techniques to the scientific community.
Technical Communication: Delivered algorithm training to diverse technical and non-technical audiences.
Framework Development: Built structured approaches for AI problem-solving using state-of-the-art transformer architectures.

Assistant Research Scientist, UMD-ESSIC / NASA-GSFC

January 2023 – March 2024

Led continental-scale forest classification using LiDAR time-series and deep learning.
Developed wildfire simulation algorithms for real-time prediction with federal partners.
Implemented hyperspectral analytics with cutting-edge neural architectures.
Maintained deployed algorithms in HPC with rigorous validation protocols.
Coordinated projects across NASA, USFS, engineers, and domain experts.

Senior Scientist & Lead Model Developer, Regrow Ag, Inc.

July 2021 – December 2022

Enhanced DNDC biogeochemical simulation model for risk and carbon cycle analysis.
Developed coupled C–N cycle models with water management & soil dynamics.
Built automation for calibration, sensitivity analysis, and validation.
Integrated satellite data via real-time assimilation pipelines.
Deployed scalable systems on AWS with Dockerized workflows.

Post-doctoral Researcher, University of Montana

April 2019 – February 2021

Developed LSTM/CNN architectures for large-area spatio-temporal prediction.
Built computer-vision models using multi-temporal satellite imagery.
Applied advanced statistics to complex environmental time-series.
Implemented distributed training & inference on HPC systems.

Graduate Research Scientist (Ph.D.), Montana State University

August 2014 – May 2019

Developed global forest dynamics in the LPJ Dynamic Global Vegetation Model.
Integrated land use, forest management, and fire dynamics into ESMs.
Created statistical methods for atmospheric CO₂ signal decomposition.
Processed terabyte-scale satellite, eddy-covariance, and atmospheric datasets.

DOMAIN EXPERTISE

Risk & Predictive Modeling

15+ years developing models for environmental risk, wildfire simulation, and agricultural systems.
Experience with real-time risk prediction and continental-scale classification.
Uncertainty quantification and temporal pattern recognition.

Regulated Environment Experience

NASA collaborations with comprehensive validation & documentation standards.
Federal partnerships requiring strict algorithm verification & testing.
Compliance-ready, reproducible systems and documentation.

Production Algorithm Development

Full lifecycle: requirements → validation → deployment → monitoring.
Failure analysis, performance optimization, and stakeholder alignment.

EDUCATION

Ph.D. Ecology & Environmental Sciences — Montana State University (2014–2019)
M.S. Environmental Sciences & GIS Certificate — Florida Atlantic University (2012–2014)
B.S. Biological Sciences — Florida Atlantic University (2006–2010)

LEADERSHIP & IMPACT

Technical Leadership

Led teams across data science, software engineering, and domain science.
Managed stakeholders across government, academia, and industry.
Established best practices for development, testing, and documentation.

Professional Activities

NASA FireSense Implementation Team (2023–2024).
USFS FASMEE Data Management Lead (2023–2024).
Talks at AGU, ESA, International Carbon Dioxide Conference.

Scientific Impact

30+ peer-reviewed publications (e.g., Nature Geoscience, PNAS, Global Change Biology).
Lead author on carbon-cycle and climate–ecosystem studies.
Open-source tools for statistical signal analysis adopted internationally.

Recognition

NASA Earth & Space Science Fellowship (2016–2019).
NSF EAPSI Fellowship (2015); JSPS Fellowship (2015).