Aaron M. Hosford

Introduction

Highly creative systems thinker and radical innovator with a deep and broad understanding of ML algorithms, theory, and applications. 18+ years of workplace experience – 30+ years counting personal projects. Track record of proven, patent-worthy designs. Seeking research or engineering role in Data Science, Machine Learning, or NLP, with a focus on novel algorithm development and/or application to new domains.

• Excellent problem-framing, problem-solving, analysis, & critical thinking.
• Strong coding, design, & data skills.
• Fast-learning autodidact and cognitive "heavy lifter".
• Grounded, focused, and clear-headed in the face of extreme domain complexity.
• Thrives on "impossible" problems.

Major Accomplishments

• Co-founder of a (non-LLM) deep-tech ML/AI startup.
• Inventor of numerous proprietary machine learning algorithms, including work with hyperdimensional computing, deep learning, genetic programming, and reinforcement learning.
• Architect of a proprietary non-LLM natural language and artificial intelligence system for open-domain, reliable, fully interpretable multi-hop reasoning and Q&A.
• Patent author for an anaphora resolution algorithm.
• ML/NLP-related contributions to the initial public release and ongoing support of a natural language financial assistant used by millions of customers of a major bank.
• Author and maintainer of the open-source XCS Python library, an object-oriented implementation of the eponymous XCS machine learning algorithm.
• National Merit Scholarship winner with full scholarship to the University of Texas at Dallas.

Summary of Technical Skills

Algorithm Design | Algorithm Development | Anaphora Resolution | Artificial Intelligence (AI) | Artificial Neural Networks (ANN) | Automated Proof Verification | Automated Theorem Provers | Automatic Differentiation | Bayesian Modeling | Business Process Automation | C | Computational Theory | CUDA | Cython | Data Annotation | Data Annotation Pipeline Design/Management | Data Mining | Data Science | Deep Learning | Dependency Parsing | Discriminative Models | Evolutionary Algorithms | eXtended Classifier Systems (XCS) | Extract‑Transform‑Load (ETL) | Generative Models | Genetic Algorithms (GA) | Genetic Programming (GP) | Graph Neural Networks (GNN) | Formal Logic | Fuzzy Logic | Hybrid ML Algorithms | Hyperdimensional Computing (HDC) | Information Theory | JSON | Knowledge Graphs | Knowledge Representation | Large Language Models (LLM) | Learning Classifier Systems (LCS) | Long Short‑Term Memory (LSTM) | Machine Learning (ML) | Mathematics | Metric Spaces | Multithreaded Applications | Natural Language Processing (NLP) | Natural Language Understanding (NLU) | Natural Language Generation (NLG) | Neural Architecture | NumPy (np) | Ontology | Optimization | Parser Design | Particle Filters | Probability Theory | Python | Reinforcement Learning (RL) | Requirements Analysis | Semantic Pointer Architecture (SPA) | SQL | Statistics | Symbolic AI | Technical Leadership | Temporal Differences (TD) | Tensor Computing | TensorFlow (tf) | Theano | Torch/PyTorch

Education

Dual major, Computer Science and Mathematics, 1997 to 2001

University of Texas at Dallas, Richardson, Texas

Awards: National Merit Scholar, Silver Anniversary Scholarship

Note: I completed the core curriculum but was forced to drop out just before graduation due to a disability-related life event. I have the education, but not the degree.

Autodidact and Machine Learning/AI Enthusiast, 1990 to current

Over 30 years of continually focused self-learning, practice, and experimentation

Relevant Experience

---

---

---

---

---