Brandon Scott

Designed and deployed production AI/ML solutions for autonomous systems using Python, PyTorch, TensorFlow, and Scikit-learn. Built scalable AI services and full-stack applications with FastAPI, Flask, TypeScript/Node.js, REST APIs, WebSockets, and microservices. Developed computer-vision and deep learning pipelines for perception and decision-making, including data preprocessing, model training, evaluation, and inference optimization. Implemented LLM, RAG, and AI agent capabilities using embeddings, vector databases, function calling, and structured outputs with leading LLM APIs. Applied MLOps practices with MLflow, Docker, Kubernetes, and CI/CD for reproducible experiments, model versioning, and production deployment. Built data-processing and evaluation pipelines using Pandas, NumPy, and SQL to support performance analysis and operational analytics.

Developed and trained machine learning and deep learning models in Python using PyTorch, TensorFlow, and Scikit-learn for classification and prediction tasks. Built data processing and feature engineering pipelines with Pandas, NumPy, and SQL, improving data quality and model performance on large datasets. Created computer-vision models for image classification and object detection, achieving approximately 90%+ validation accuracy across selected use cases. Performed model evaluation and optimization to improve accuracy and reduce inference time, comparing algorithms and hyperparameters. Built ML workflows using MLflow, Docker, Git, and cloud services for experiment tracking, model versioning, and deployment. Created Python APIs and internal tools to integrate models with software systems and deliver prediction results to downstream systems.

Developed and maintained backend services with Java, Python, Spring Boot, and REST APIs, improving application response time and performance. Designed and integrated RESTful APIs for internal and customer-facing applications, reducing API response latency. Optimized SQL queries and database operations to reduce query execution time. Built and optimized containerized microservices using Docker and Kubernetes, improving deployment consistency and reducing deployment time. Created unit and integration tests and contributed to Jenkins/Git-based CI/CD pipelines to reduce manual testing and deployment effort. Investigated production issues using logs and performance data, implementing fixes that improved reliability and reduced recurring incidents.