Lead data-driven initiatives focused on operational efficiency, mass and energy balance validation, and biomass-to-energy conversion performance across industrial steam generation units. Develop analytical frameworks to diagnose process inefficiencies, validate measurement systems, identify operational losses, and support tactical and operational decision-making through statistical modeling, industrial analytics, and data engineering. Responsibilities include analyzing inconsistencies between steam production, billing, and customer measurement systems; detecting mass and energy losses across different operational regimes; modeling operational efficiency and biomass consumption drivers; and developing predictive and anomaly detection models for steam demand, production behavior, and efficiency deviations. Develop analytical dashboards, operational indicators, benchmarking frameworks across industrial units, automated calculation pipelines, and operational alert systems to support KPI monitoring, efficiency improvement, cost decomposition analyses, and loss reduction initiatives. Build analytical solutions integrating the full operational chain (biomass moisture, calorific value (PCI), combustion efficiency, steam generation, thermal losses, and operational costs), translating industrial process data into measurable operational and financial insights.

I worked as a Data Scientist supporting Sugarcane R&D at CTC (Centro de Tecnologia Canavieira), where I led the application of advanced statistical and experimental methodologies across large-scale field trials and research programs, enabling data-driven decision-making in genetic improvement, biotechnology development, and seed production systems. My work focused on multi-location and multi-year field trials, addressing complex experimental structures with hierarchical and unbalanced designs and enabling the evaluation of genotypes across diverse environments. I developed and deployed advanced statistical modeling frameworks, including linear and generalized linear mixed models and extensions for non-normal responses and heterogeneous variance structures, driving the modeling of disease resistance, yield components, and key agronomic traits and translating outputs into decision-oriented insights for breeders and agronomy teams. In parallel, I established reproducible, production-grade analytical pipelines spanning data processing, model development, and result synthesis, and designed analytical dashboards and reporting layers to operationalize insights, support real-time monitoring of key metrics, and directly inform strategic and operational decision-making across R&D programs.

I worked as a Data Scientist in the Forests, Land Use, and Agriculture Program at WRI Brasil, where I developed and implemented data engineering solutions within the Data Management and Monitoring team. My work focused on structuring, processing, and analyzing complex datasets, including both geospatial and tabular data, to develop tools and methodologies that supported data-driven decision-making in forest restoration, sustainable land use, and environmental monitoring. I was responsible for designing and maintaining APIs and ETL pipelines optimized for geospatial data processing. This included developing spatial models, performing advanced analyses, and optimizing PostgreSQL/PostGIS queries to improve performance and scalability. I worked with multiple government data sources, including SICAR, MMA, FUNAI, PRODES, Global Forest Watch (GFW), SIGEF/INCRA, and IBGE, ensuring proper integration, standardization, and effective use of these datasets. In addition, I developed scalable REST APIs to enable real-time access and integration of geospatial data across different platforms. I built geospatial solutions using GeoServer, GIS technologies, and Python, leveraging libraries such as GeoPandas, Shapely, Rasterio, and Fiona to design data pipelines, perform spatial analyses, and support environmental monitoring and management initiatives.

Worked as a Data Science Researcher at Fiocruz Ceará, contributing to epidemiological surveillance and genomic monitoring initiatives focused on high-impact pathogens affecting the Brazilian population. Developed analytical pipelines and predictive modeling solutions for genomic data analysis and outbreak monitoring, supporting data-driven public health surveillance and epidemiological research. Integrated statistical and machine learning models with genomic and epidemiological datasets to support outbreak prediction, surveillance workflows, and scientific analysis based on real-time and large-scale health data.

Worked as a Data Analyst at Doce Barreira, leading analytical and automation initiatives to support operational management, business performance monitoring, and administrative decision-making across commercial operations. Managed end-to-end data workflows involving data structuring, integration, analysis, and reporting to improve operational visibility and business process efficiency. Designed and implemented centralized data management and reporting solutions integrating inventory, sales, and cash flow information, enabling real-time monitoring of key business metrics and improving data consistency across operational processes. Developed automation tools and analytical workflows using Python, Bash, Excel/VBA, and Linux-based environments to reduce manual workloads, streamline administrative routines, and support scalable business operations.