Aaron M. Hosford

903-327-2785 | hosford42@gmail.com | 75071

Website | GitHub | HTML Resume | PDF Resume | LinkedIn | Patent Info

Seeking: Senior ML Researcher, Senior Data Scientist, Senior ML Engineer

Expertise: ML Model Design, Algorithm Design, NLP, NLU, NLG, Software Dev


Introduction

Highly creative systems thinker and radical innovator with a deep 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 senior research or engineering role in Data Science, Machine Learning, or NLP, with a focus on novel algorithm development and/or application to new domains.


Major Accomplishments


Summary of Technical Skills

Accuracy-based Classifier Systems (XCS) | Algorithm Design | Algorithm Development | Anaphora Resolution | Array Computing | Artificial Intelligence (AI) | Artificial Neural Networks (ANN) | Automated Proof Verification | Automated Theorem Provers | Automatic Differentiation | AWS SageMaker | AWS S3 | Bayesian Modeling | Business Process Automation | C | Classification | Computational Theory | Convolutional Neural Networks (CNN) | CUDA | Cython | Data Annotation | Data Annotation Pipeline Design/Management | Data Mining | Data Science | Database Design | Deep Learning | Dependency Parsing | Discriminative Models | Entity Detection | Evolutionary Algorithms | Excel | Extract-Transform-Load (ETL) | Generative Models | Genetic Algorithms (GA) | Genetic Programming (GP) | Graph Neural Networks (GNN) | Flask | Formal Logic | Fuzzy Logic | Hyper-Vector Computing (HVC) | Information Theory | Intent Detection | Java | JSON | Keras | Knowledge Graphs | Knowledge Representation | Large Language Models (LLM) | Learning Classifier Systems (LCS) | Lisp | Long Short-Term Memory (LSTM) | Lua | Machine Learning (ML) | Mathematics | Metric Spaces | Multithreaded Applications | Natural Language Processing (NLP) | Natural Language Understanding (NLU) | Natural Language Generation (NLG) | NumPy (np) | Ontology | Optimization | Pandas (pd) | Parser Design | Particle Filters | Pillow/PIL | Probability Theory | Python | R | Recurrent Neural Networks (RNN) | Regression | Reinforcement Learning (RL) | Reinforcement Learning from Human Feedback (RLHF) | Requirements Analysis | SciKit-Learn | SciPy | Scripting | Semantic Pointers | SpaCy | SQL | Statistics | Supervised Learning | Symbolic AI | Temporal Differences (TD) | Tensor Computing | TensorFlow (tf) | TensorFlow Probability (tfp) | Theano | Unsupervised Learning

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 focused self-learning, practice, and experimentation


Relevant Experience

Role Co-Founder and Chief Technology Officer
Employer Transparent AI
Location Remote
Duration Feb 2024 to current
Skills Leadership, Algorithm Design, Product Design, Requirements Analysis, ML, Programming
Summary CTO and co-founder of a startup in the Explainable and Auditable AI space.
Accomplishments • Developed bespoke algorithms for explainable AI.
• Helped design the product from the ground up.
• Managed and worked closely with product development team to rapidly develop a working prototype.
• Assisted in pitching directly to investors.

Senior Software and Machine Learning Engineer
Employer Wiser Solutions, Inc.
Location Remote
Duration Nov 2022 to Feb 2024
Skills ML, Image Processing, Algorithm Design, Requirements Analysis, Leadership
Summary Helped develop a product matching pipeline for online retail pricing.
Accomplishments • Developed a bespoke high-throughput image clustering algorithm for reliably identifying representative images for similar products across multiple online retailer sites.
• Designed & developed a data labeling pipeline for crowdsourcing training and evaluation samples appropriate to the problem domain.
• Comprehensive model analysis and evaluation, including the development of a novel framework for evaluation metrics appropriate to the novel problem space of N-ary representative image selection.
• Designed a responsive pipeline for continually applying the representative image selection algorithm to a live data stream.

Data Scientist
Employer Insight Global (initial contract), Bank of America (hired full-time)
Location Plano, TX
Duration Oct 2017 to Nov 2022
Skills NLP, ML, Data Science, Algorithm Design, Programming, Requirements Analysis, Leadership
Summary Conversational agent: Designed, implemented, tested, and evaluated entity recognition, intent detection, and anaphora resolution models & algorithms.
Accomplishments • Patent pending for an anaphora resolution algorithm.
• Part of the team responsible for the ML and NLP subsystem of a virtual financial assistant – a groundbreaking chatbot serving tens of millions of bank customers .
• Significant improvements to intent and entity recognition model performance.
• Extensive enhancements to the team’s data handling processes and model metrics.
• Built natural language annotation tools from scratch.
• Built bespoke heuristic/rule-based system for time entity extraction.
• Regularly collaborated with other experts to improve customer experience in accordance with stakeholder direction.

Role Lead Systems Architect
Employer Ericsson
Location Plano, TX
Duration Jun 2014 to Sep 2017
Skills Programming, ML, Algorithm Design, Requirements Analysis, Leadership
Summary Projects ranging over a variety of domains, including business process automation, explainable AI applications, and supply chain forecasting pipeline.
Accomplishments • Trained & led a remote team in an agile environment.
• Designed and built an automated supply chain forecasting pipeline.
• Worked closely with stakeholders to identify business requirements.
• Evaluated and applied a multitude of ML algorithms, including deep learning, decision trees, market basket analysis, and ARIMA.
• Designed and developed an open-sourced Python business automation framework.

Programmer/Software Analyst
Employer West Asset Management (now Alorica)
Location Sherman, TX
Duration Nov 2006 to Jan 2014
Skills Programming, Automation Design, Requirements Analysis, ETL, Leadership
Summary Business process automation for varied domains including ETL, scripting, web scraping, and financial reporting.
Accomplishments • End-to-end design, implementation, and automation of client-facing financial and account data interfaces, handling financial data in excess of $50,000,000+ with associated personally identifiable information and health data.
• Conception, design, and coding of extensive contributions to the in-house business automation library.
• Regularly relied on for reviews and difficult debugging scenarios for code running on remote systems.

Software Engineer (Intern)
Employer Ericsson
Location Sherman, TX
Duration Aug 1998 to Nov 2001
Skills Programming, OPNET Simulation, Product Design
Summary Undergrad working as a graduate-level software engineering intern for a collaborative project between Ericsson and UT Dallas to research VoIP telephony quality.
Accomplishments • Coded network traffic analysis simulations.
• Built a telephony switch documentation management system.