Daniel Crawford

• Led end-to-end development and deployment of deep reinforcement learning models for autonomous underwater vehicles (AUVs), including designing and building a custom simulation environment to validate agent performance in realistic scenarios • Collaborated on the development of a vendor contract matching system using Large Language Models (LLMs), custom APIs, and full-stack software tools to automate and improve contract analysis workflows • Led the design and deployment of an OCR/Text Extraction pipeline to extract structured data from PDFs, enhancing downstream LLM and ML matching capabilities • Contributed to building a Retrieval-Augmented Generation (RAG)-based search engine to identify foreign company ownership and associated risk factors through automated web crawling and smart entity linking • Engaged in engineering/developer culture through code review, emphasis on clean, SOLID code

• Designed and deployed time-series and classification machine learning models to predict real-time events in live sporting matches • developed API deployment infrastructure on GCP used to host machine learning models, improving scalability and enabling new model types to be able to be deployed • data visualization and strategic analysis meetings using Python, driving actionable insights that improved decision-making for model selection type

• Designed and Built text summarization software to parse resumes as lead engineer • Created Machine Learning/LLM Pipelines that founded the organization-wide basis for product. • Integrated Mongo DB and OpenAI ChatGPT, Hugging Face, and LLAMA API to automate training and testing data • Established database connections and engineered text data pipelines decreasing training time • Leveraged real-world business expertise and needs in conjunction with theoretical NLP and Linguistics

• Discovered over $XXM in client profit gains using a cutting-edge AI prescriptive analysis model. • customer-facing solutions engineering data science/engineering initiatives, ensuring alignment with client goals and accelerating solution delivery, including automated machine learning • Boosted customer acquisition through optimized data integration and analytics, providing valuable insights that guided targeted customer engagement strategies, decreasing time-to-delivery of analytics output

• Built automated data ingestion pipeline for a municipal energy company • Formatted data pipeline in Snowflake’s elastic data warehouse • Synchronized Snowflake warehouse with AWS infrastructure with Terraform • Provided SQL queries from SQL server to Analytics Dashboard • Tested automated data pipeline, built with CI/CD cycle

• Led data science effort for the creation of metrics and discrete event simulation software • Built Automated Analytics library in R to analyze sustainability operations • Designed and built R Shiny dashboards for specific analysis types • Created a research article search engine with NLP summarization techniques • Developed analytics dashboard in Java for Army General Officers to predict and apportion budget • Mentored junior Data Science/ORSA Interns and instructed in data science and project management • Led data collection from cross-disciplinary soldier and civilian specialist teams • Received award from SES-2, institute director, in recognition for leadership of novel data science effort • Recipient of DoD Science, Math, and Research for Transformation (SMART) Scholarship

Pitt's Bachelor of Philosophy Program Independent and novel research in a cross-discipline project. My thesis Language Family and Geocomputation can be found here: http://d-scholarship.pitt.edu/38738/

Pitt's Brackenridge Fellowship for Multi-Disciplinary Projects Conducted research in Computer Science/Mathematical Modeling, and Psychology