George
Cubas
Machine Learning Engineer
Machine Learning Engineer with 10+ years of combined experience in tech and energy. I build scalable AI solutions that drive revenue — from oil & gas economics to fintech analytics and real estate valuation models.
Skills & Expertise
Deep specialization in Python/ML with strong engineering fundamentals spanning backend systems, cloud infrastructure, and energy domain expertise.
ML & Data Science
Backend & Cloud
Engineering & Domain
Work Methodology
How I approach building ML systems that deliver measurable business impact.
Discovery & Analysis
Understanding business objectives and data landscapes before writing a single line of code.
Scalable Architecture
Building modular, maintainable ML pipelines and APIs designed to handle production traffic.
Rapid Iteration
Fast experiment cycles with clear metrics, moving from prototype to production efficiently.
Continuous Optimization
Performance tuning, model retraining, and infrastructure improvements as standard practice.
Work Experience
A decade of building ML solutions across energy, fintech, real estate, and enterprise — each role compounding domain expertise with engineering depth.
Python Developer
- Deployed production-grade FastAPI microservice for Stripe payment processing
- Migrated invoice generation from Pandas to Polars for parallelized data pipelines
- Built Polars-powered reporting engine for structured monthly financial summaries
- Implemented async webhook handlers with PostgreSQL via SQLAlchemy ORM
Python Developer
- Developed Oil & Gas economic models for cost recovery and profit split analysis
- Created DDA functions enhancing cash flow models with Polars parallelization
- Implemented NPVI and Profitability Index calculations for accurate cash flow analysis
- Designed FastAPI endpoints for advanced scenario sensitivity analysis
Python Developer
- Built RAG system using Dolphin LLaMA with Dspy and Weaviate vector DB
- Designed AI services including image classifiers using natural evolution methods
- Maintained CI/CD pipelines with test-driven development practices
Python Developer
- Engineered ETL pipelines for Office of Sponsored Programs using Pandas & SQLAlchemy
- Built web crawlers for frontend data quality verification
- Leveraged multiprocessing and asyncio for parallelized performance
Python Developer
- Built ML valuation models using Pandas, NumPy and Scikit-learn
- Scraped Freddie Mac & Case Shiller data with BeautifulSoup & Selenium
- Applied Monte Carlo Analysis for Levered and Unlevered IRR prediction
Python Engineer
- Built Anti-Collision Risk Analysis app using Django Framework
- Created neural network-based document extraction pipeline
- Developed ML models for production drawdown using Scikit-learn
Drilling Engineer
- Managed drilling operations on the North Slope of Alaska
- Foundation in energy sector operations and engineering economics
Projects
Open-source projects showcasing ML engineering, distributed systems, and applied AI research.
RAG Pipeline
Generative AI with Weaviate & Dspy
A production RAG framework integrating multiple data sources with Weaviate vector DB and Dspy for enriching LLM knowledge bases and enhancing contextual accuracy.
Impact: Multi-source retrieval with context-aware generation
Apache Spark Prediction Pipeline
IoT Data & Pressure Prediction
End-to-end Spark pipeline using IoT sensor data for real-time pressure prediction with distributed computing for high-throughput data processing.
Impact: Real-time prediction at scale with IoT data
Reinforcement Learning NN
Actor-Critic for CartPole
An Actor-Critic reinforcement learning algorithm that uses an actor network for optimal policy discovery and a critic network for action probability evaluation to solve the CartPole problem.
Impact: Convergent policy with optimized reward function
Let's build something impactful together
Full-Time / Contract / Consulting — available for ML engineering, data pipeline architecture, and AI product development.