Randley Morales, Ph.D.

Machine Learning Engineer | Generative AI | Computer Vision

PROFESSIONAL SUMMARY

I am Randley Morales

Ph.D. Mathematician & AI Engineer with 10+ years of experience bridging advanced theory and production-grade ML. Specialized in Generative AI, Computer Vision, and RAG architectures, I leverage a deep background in Differential Geometry to build high-performance, scalable models. From research to API deployment, I transform complex high-dimensional data into measurable business impact.

TECHNICAL SKILLS

  • Generative AI & LLMs: RAG Architectures, LangChain, Semantic Search, Vector Databases (Pinecone, ChromaDB), Prompt Engineering, OpenAI API, Hugging Face.
  • Machine Learning: XGBoost, Random Forests, Gradient Boosting, PCA, K-Means, Hyperparameter Tuning (GridSearch, Optuna), A/B Testing.
  • Deep Learning & Computer Vision: CNNs (ResNet, VGG, EfficientNet), Object Detection (YOLO), Image Segmentation, Transfer Learning, OpenCV.
  • Programming & MLOps: Python (Pandas, NumPy, Scikit-learn), PyTorch, TensorFlow/Keras, SQL, Docker, Flask, Streamlit, Git.
  • Mathematics: Differential Geometry, Computational Geometry, Numerical Linear Algebra, Differential Equations, Probability Theory, Statistics.

SuperKart

| End-to-End Sales Forecasting Engine

Engineered a full-stack revenue forecasting system to support quarterly planning, scenario testing, and stakeholder decision-making.

Daily anti-aging cream

HelmNet

| Helmet Detection for Industrial Safety

Developed a computer vision model to detect safety helmets in high-risk environments, designed for real-world lighting and viewpoint variation.

Medical Assistant

| Generative AI & RAG Pipeline

Built a Retrieval-Augmented Generation (RAG) system to synthesize medical information from unstructured text with high-precision semantic retrieval and domain-aware responses.

ReneWind

| Predictive Maintenance for Wind Turbines

Built a failure prediction model on imbalanced sensor data to proactively identify components at risk and reduce downtime.

EasyVisa

| Visa Approval Prediction Model

Developed an ensemble classification model to predict visa approval outcomes using demographic and economic indicators.

Personal Loan Campaign

| Predictive Customer Classification

Developed a machine learning classification model to identify customers most likely to accept personal loan offers, enabling more targeted and efficient marketing campaigns.

FoodHub

| Data-Driven Revenue Strategy

Performed comprehensive exploratory data analysis (EDA) on restaurant order data to identify demand patterns, customer preferences, and revenue opportunities across cuisines.

CURRICULUM VITAE

Explore my complete CV, experience, and technical background

View my full professional background in mathematics, machine learning, generative AI, computer vision, predictive modeling, research, and teaching. My CV includes academic training, industry experience, technical skills, certifications, and applied AI projects from research to deployment.

Where Mathematics meets Machine Learning and Photography.  

Where Mathematics meets Machine Learning and Photography.  

Where Mathematics meets Machine Learning and Photography.  

Where Mathematics meets Machine Learning and Photography.  

Where Mathematics meets Machine Learning and Photography.  

Through patterns, geometry, light, and data, I study how visual structures emerge and how algorithms can learn from them. My work connects artistic exploration with mathematical thinking to better understand images and visual intelligence. Photography.
PORTFOLIO

Building intelligent systems with mathematics, machine learning, and real-world data.

View All Projects
Machine learning sales forecasting dashboard for retail analytics and predictive modeling developed in the SuperKart forecasting engine by Randley Morales
SuperKart | Machine Learning Sales Forecasting EngineArtificial Intelligence and Machine Learning

SuperKart | Machine Learning Sales Forecasting Engine

SuperKart is a machine learning sales forecasting system built with Python to predict retail demand using historical data and predictive analytics. The project uses feature engineering and gradient boosting models to generate accurate forecasts and support data-driven retail planning.
Computer vision helmet detection system for industrial safety using deep learning developed in the HelmNet machine learning project by Randley Morales
HelmNet | Computer Vision for Industrial SafetyArtificial Intelligence and Machine Learning

HelmNet | Computer Vision for Industrial Safety

HelmNet is a computer vision deep learning project designed to detect safety helmets in industrial environments. The system uses convolutional neural networks to identify protective equipment in images and support automated workplace safety monitoring.
AI Medical Assistant using Generative AI and RAG pipeline for healthcare question answering
Medical Assistant | Generative AI & RAG PipelineArtificial Intelligence and Machine Learning

Medical Assistant | Generative AI & RAG Pipeline

Medical Assistant is an AI-powered healthcare assistant built using natural language processing, large language models, and retrieval-augmented generation (RAG). The system retrieves relevant medical knowledge and generates accurate responses to support intelligent medical question answering.
ReneWind machine learning predictive maintenance system for wind turbines in renewable energy
ReneWind | Predictive Maintenance for Renewable EnergyArtificial Intelligence and Machine Learning

ReneWind | Predictive Maintenance for Renewable Energy

ReneWind is a machine learning predictive maintenance system for wind turbines in renewable energy infrastructure. The project analyzes turbine operational data, builds predictive features, and trains models to identify failure risks and optimize maintenance planning.

Machine Learning | Randley Morales | Deep Learning | 

Machine Learning | Randley Morales | Deep Learning | 

Machine Learning | Randley Morales | Deep Learning | 

Machine Learning 

Machine Learning 

Machine Learning 

Machine learning and deep learning predictive model visualization showing artificial intelligence data analysis by AI engineer Randley Morales