Machine Learning Data Scientist

Bryan Bannayan Aval

Email address: **********@*****.***, **********@*****.***

Linkedin: https://www.linkedin.com/in/bannayan/

Aliso Viejo, California, USA

Summary

Ph.D. in Environmental Physics, Data Professional with 12+ years of experience building scalable data platforms and deploying machine learning solutions across climate, agriculture, and manufacturing domains. Highly accurate and experienced Data Scientist adept at collecting, analyzing, and skilled at transforming complex, high-volume datasets into actionable insights through automated ETL, predictive modeling, time-series analysis and employing any required Statistical analysis. Preparing reports and oral presentations about data in industry. Possessing an extensive analytical skills, strong attention to detail, and a significant ability to work in team environments. Developing machine learning (XGBoost, KNN, Random Forest and logistic regression) and biophysical models and proficient in optimizing ML and NN models performance to enhance predictive accuracy and operational efficiency. Skills

Programming & Tools: Python, R, SQL, GCP (BigQuery), AWS, Domino, Snowflake, MLflow, Workflow Automation Tools, PowerBI,

Modeling: Machine learning (ML), Deep learning (DL), Convolutional Neural Networks

(CNN), Model evaluation, Model Optimization, Error analysis, Hyperparameter tuning

Data Science & Analysis: Data preprocessing, exploratory data analysis (EDA), quality control/assurance (QC/QA), statistical analysis

AI Workflow Automation & Hybrid architecture: AI agent development and process orchestration, Retrieval-Augmented Generation (RAG): Design, and implementation for domain-specific question answering and knowledge retrieval.

Scientific Communication: Academic writing, scientific publication, data interpretation, research reporting

Analytical & Problem Solving: Strong attention to detail, critical thinking, data-driven decision-making

Leadership & Collaboration: Team leadership, mentoring, business intelligence, effective communication

Project & Time Management: Proven ability to manage multiple projects and meet deadlines

Professional history

Bayer, Senior Data Scientist, Berkeley CA (Remote), Aug 2021-Jun 2025

Built and deployed data-driven models (Cool Farm Tool, DSSAT, and ML) to quantify GHG emissions and soil carbon trends under diverse scenarios.

Directed a four-member data science team in designing climate adaptation pipelines, leveraging predictive modeling to optimize performance of Bayer’s products under future climate and water constraints.

Applied advanced optimization techniques to balance nitrogen fertilizer use across yield, emissions, and economic trade-offs.

Engineered scalable, reproducible Python pipelines for global GHG datasets and developed cloud-based DSSAT workflows on GCP with BigQuery integration and Vertex AI.

Enhanced predictive accuracy of disease forecasting models through ML, delivering real- time, cloud-native simulations and minimizing local computation overhead.

Authored technical documentation, reports, and presentations, translating complex modeling insights into business-aligned, actionable strategies.

Drove Bayer’s sustainability goals by delivering advanced data science solutions that improved productivity while reducing environmental impact. ITERIS, Data Scientist, Berkeley CA (Remote), Jun 2018-Jun 2021

Developed predictive models to quantify disease risks, environmental stress, and climate– crop interactions, leveraging large-scale datasets and validated with peer-reviewed research.

Developed scalable feature engineering workflows (one-hot encoding, PCA) with experiment tracking in Git, leveraging ANN and SVR models to enhance prediction accuracy for wheat moisture content and crop yields.

Designed mathematical algorithms and neural networks using weather and environmental datasets, with results published in studies on climate variability, cereal yield forecasting, and productivity under elevated CO and temperature.

Built automated, scalable pipelines for irrigation estimation, weather/soil data processing, and integration into decision-support tools, advancing optimization of resource use and climate adaptation strategies.

Enhanced national-scale climate risk models and streamlined irrigation workflows by combining data engineering with predictive analytics, validated against real-world datasets and published methodologies.

Advanced modeling frameworks by integrating field data, statistical methods, and scenario-based simulations, with demonstrated impact in publications on water stress, yield forecasting, and climate resilience strategies.

Washington State University, Data Modeling, Prosser Campus WA, Aug 2017-May 2018

Applied advanced statistical and machine learning techniques to model drought, climate variability, and risk assessment at regional and national scales.

Translated complex simulation outputs into actionable insights for product teams, stakeholders, and decision-makers, enabling data-informed adaptation strategies such as optimized planting schedules.

Directed predictive analytics projects on climate risk and genotype–environment interactions, improving yield forecasting, resource optimization, and management strategies.

Built scalable data pipelines and forecasting tools, including KNN-based weather analogue models, daily temperature forecasting workflows, and large-scale spatio- temporal data integration systems.

Developed data-driven frameworks to identify optimal environments for specific genotypes and supported regional decision-making with precision forecasting and risk assessment tools.

Authored 80+ peer-reviewed publications, demonstrating expertise in data-driven modeling, predictive analytics, and the application of machine learning to complex environmental systems.

The University of Georgia, Data Modeling, Griffin, GA, Jul 2015-Jul 2017

Applied advanced statistical and machine learning techniques to analyze large-scale field and experimental datasets, supporting model calibration, validation, and productivity assessments under variable conditions.

Quantified interactive effects of weather, soil, and management factors on productivity using predictive analytics and ML algorithms, enabling evaluation of genotype performance across diverse environments.

Designed and implemented predictive tools and models combining simulation, statistical methods, and data-driven approaches to optimize performance outcomes.

Built scalable data pipelines for climate–crop modeling, integrating spatio-temporal datasets to support irrigation analysis, yield forecasting, and adaptation strategies.

Delivered data-driven insights for genotype–environment optimization, providing precision forecasting and regional decision support for climate resilience and resource efficiency.

Advanced modeling and analytics expertise through research on drought vulnerability, climatic indices, and nitrogen use efficiency, demonstrating ability to translate complex datasets into actionable decision-support systems. Education

PhD, University of Nottingham, UK, Environmental physics

“Torture the data, and it will confess to anything.” – Ronald Coase

Contact this candidate