Experience
Independent AI Research - Applied AI Engineer
July 2024 - Invalid Date
- Developed Alfred, a Python-based Model Context Protocol (MCP) server that exposes enterprise APIs to LLM engines.
- Implemented a decoupled SourceAdapter interface to synchronize information and reinforce infrastructure scalability.
- Evaluated deep reinforcement learning policies using Ray RLlib and SimulationCraft to optimize agent decision-making logic.
- Benchmarked reinforcement learning policies against deterministic baseline strategies to measure performance.
- Fine-tuned Qwen2.5-1.5B-Instruct using 4-bit QLoRA on a single 8GB GPU for a specialized persona-adaptation task.
- Reduced held-out model perplexity by 54% through the implementation of completion-only loss masking.
- Architected an automated RAG pipeline to ingest, index, and optimize text-chunking for software documentation.
- Designed semantic search strategies to maximize retrieval precision and reduce downstream LLM hallucination.
Pennsylvania State University - Research Assistant
December 2022 - July 2024
- Engineered an automated multi-class text classification pipeline using Python, SQL, and commercial LLM APIs.
- Mapped 37 years of historical research corpora across 6,607 documents using LLM classification pipelines.
- Developed custom regex, tokenization, stop-word filtering, and lemmatization pipelines in SpaCy and NLTK.
- Cleaned raw text strings to optimize computational efficiency and token context-window density.
- Implemented Latent Dirichlet Allocation (LDA) modeling with asymmetric alpha priors to extract latent word-over-topic distributions.
- Validated and iterated a 15-tier methodological taxonomy using unsupervised topic modeling.
- Architected validation frameworks to benchmark LLM accuracy against a human-annotated ground truth matrix.
- Optimized classification confidence via token-level logprob thresholding.
- Calculated quantitative validation metrics including the Jaccard Index, precision, recall, and F1 scores.
Intel - Automation Quality Analyst
July 2021 - November 2022
- Reviewed and analyzed automated test execution logs daily to isolate root causes of failures.
- Distinguished accurately between true system regressions and internal automation framework errors.
- Summarized log findings and escalated actionable technical details to relevant domain managers.
- Communicated cross-functionally across domain groups to ensure swift resolution of development blockers.
School of Civil and Construction Engineering at Oregon State University - Industry Liaison Assistant
August 2018 - August 2019
- Compile, calculate, and visualize school graduation and employment statistics.
- Maintain a directory of industry contacts in Salesforce.
- Support the organization and execution of events with industry partners.
- Automate frequent tasks with Python scripts.
Top Deck Hobbies - Store Manager
May 2014 - May 2016
- Cultivated an efficient shipping team through leadership and coworking.
- Ship an average of 600 online orders of trading-card game cards daily.
- Maintain relationships with and place orders to product distributors.
- Cultivated an efficient shipping team through leadership and coworking.
- Operate in-store weekly and regional trading-card game tournaments.
- Travel the West Coast of the United States to represent the company at Magic: the Gathering tournaments by hosting a booth to buy and sell trading cards.
Selected Courses
CMPSC 448:
Machine Learning and Artificial Intelligence
•
IST 504:
Foundations of Theories and Methods of Information Sciences and Technology Research
•
IST 505:
Foundations of Theories and Methods of Information Sciences and Technology Research
•
IST 557:
Data Mining: Techniques and Applications
•
IST 558:
Data Mining II
•
IST 597:
Foundations of Deep Learning
•
STAT 500:
Applied Statistics
•
CS 325:
Analysis of Algorithms
•
CS 344:
Software Engineering I
•
CS 362:
Software Engineering II
•
MTH 306:
Matrix & Power Series Methods
