Kebson St Charles
AI/ML Engineer | Data Scientist | Software Developer
Machine Learning
Deep Learning
Data Science
NLP
Computer Vision
AI/ML Engineer | Data Scientist | Software Developer
I'm a passionate AI/ML Engineer and Data Scientist with a unique blend of military discipline, software engineering expertise, and cutting-edge AI knowledge. Currently serving as a Software Engineer II while actively transitioning into the AI field.
My journey spans from military service to homeland security, and now to the exciting world of artificial intelligence. I hold a BS in Computer Science and an MS in Data Analytics, and have completed an AI/ML Bootcamp at Caltech, where I mastered the latest techniques in machine learning, deep learning, and AI applications.
I specialize in building intelligent systems that solve real-world problems, from predictive analytics that improve government operations to conversational AI that enhances user experiences. My projects demonstrate proficiency in Python, TensorFlow, NLP, computer vision, and cloud-based AI solutions.
What drives me: The intersection of data, intelligence, and impact. I'm passionate about leveraging AI to create solutions that matter, whether it's optimizing city operations, automating complex workflows, or building the next generation of intelligent applications.
Leverage proven APIs and services when:
Examples: OpenAI for conversational AI, Azure for document processing
Build proprietary models when:
Examples: Municipal permit prediction, specialized document classification
Combine both strategies to:
My Approach: Start with third-party, customize where differentiation matters
Every AI implementation decision is driven by three key factors: Business Value, Cost Efficiency, and Technical Feasibility. I analyze these factors to recommend the optimal approach for each specific use case.
Showcasing practical applications of artificial intelligence and machine learning
I leverage both third-party AI solutions and custom model development based on business needs, cost-effectiveness, and functional requirements. My approach prioritizes rapid deployment with proven APIs while building custom solutions when competitive advantage or specific functionality demands it.
Developed a hybrid AI system combining Azure AI Services for document processing with custom ML models for permit prediction. Strategic decision: Used Azure Cognitive Services for OCR (cost-effective, proven accuracy) while building custom ensemble models for domain-specific predictions.
Impact: 35% efficiency improvement, 90% prediction accuracy
Strategy: Third-party for commoditized tasks, custom for competitive advantage
Built intelligent chatbots using OpenAI GPT-4 and Claude AI APIs for natural conversation, while integrating Azure Copilot for code generation. Strategic choice: Leveraged best-in-class third-party models rather than training custom LLMs due to cost and complexity.
Impact: 85%+ user satisfaction, 60% faster development cycles
Strategy: API-first approach for LLMs, custom logic for business workflows
Created ML-powered dashboards using Power BI's built-in AI features combined with custom Python models. Strategic approach: Used Power BI's AutoML for rapid prototyping, then developed custom time series models for better domain-specific accuracy.
Impact: 90% prediction accuracy, faster business insights
Strategy: Rapid prototyping with built-in tools, custom models for precision
Integrated Cursor AI and GitHub Copilot into development workflows, while building custom code review and quality assurance models. Cost analysis showed 70% productivity gains justified premium tool costs vs. building internal code generation.
Impact: 70% faster development, improved code quality
Strategy: Premium tools for core productivity, custom models for quality control
Implemented Azure Document Intelligence for standard document types, with custom computer vision models for specialized municipal forms. Hybrid approach balanced cost (Azure's per-page pricing) with accuracy for unique document types.
Impact: 40% reduction in manual processing, 95% accuracy
Strategy: Third-party for standard docs, custom models for specialized forms
Combined ChatGPT API for conversational abilities with custom educational content models and Alexa Skills. Business case: Leveraged OpenAI's conversational AI while maintaining control over educational content and learning paths.
Impact: Enhanced student engagement, accessible learning
Strategy: API for conversation, custom models for educational expertise
City of Weston, FL | September 2024 - Present
August 2024 - Present
City of Weston, FL | May 2021 - September 2024
Nova Southeastern University, Davie, FL | December 2018 - December 2020
Transportation Security Administration (TSA) | 2010 - 2014
U.S. Army | Army Reserve (42A) & Active Duty (19K) | 2001 - 2010
AI/ML Bootcamp (Completed)
Advanced coursework in:
Master of Science, Data Analytics
Specialized in:
Bachelor of Science in Computer Science
Strong Foundation in Core Computer Science Principles
Key coursework relevant to AI/ML:
Continuous learning and professional development in AI/ML technologies
California Institute of Technology
Advanced ML algorithms & model deploymentCalifornia Institute of Technology
Neural networks & deep learning architecturesCalifornia Institute of Technology
Statistical analysis & predictive modelingCalifornia Institute of Technology
Python, TensorFlow, PyTorch implementationStaying current with rapidly evolving AI/ML technologies through ongoing education, hands-on projects, and industry best practices. Currently expanding expertise in generative AI, large language models, and advanced MLOps practices.
Ready to discuss AI/ML opportunities or collaborate on innovative projects?
Reach out at kebsonstcharles@gmail.com