Currently an Extern @Pfizer
Currently an Extern @PfizerBuilding hands-on experience in AI-powered document insights through a Pfizer externship while completing my degree.
Projects building applications that make use of AI algorithms, techniques, and systems.
Pfizer · ✅ Verified by Extern · ⏱️ In progress
Prototype AI-powered document intelligence with Pfizer—using OCR, LLMs, and RAG to automate real enterprise PDF workflows and build a standout portfolio project.
Kaggle Competition · Aug 2026
Built a custom reinforcement learning agent to play Pokémon TCG using a simulated game environment and PyTorch. Designed reward shaping for card sequencing, board development, energy management, damage, knockouts, and win conditions. Integrated multiprocessing to run parallel battle environments and
Personal Project · May 2026
Engineered a real-time edge AI health-screening system on a Raspberry Pi 5, integrating YOLO (Ultralytics) for automated person detection and tracking in public environments. Trained and deployed a logistic regression classifier on multi-sensor physiological data, achieving 94% accuracy for early il
Personal Project · Feb 2026
Fine-tuned Stable Diffusion (v1.5/XL) on a curated high-resolution image dataset using DreamBooth and LoRA for high-fidelity subject consistency. Deployed a scalable generative inference pipeline on Hugging Face Spaces with a Gradio interface for real-time, user-prompted image generation. Reduced in
Personal Project · Dec 2025
Collaborated with a classmate to build a solution for the land-development industry: an ensemble classifier predicting buildable lot counts from a given plot. Combined multi-modal OCR, packing algorithms, and a CNN for plot-shape analysis; identified class imbalances and proposed data-centric fixes.
Personal Project · Oct 2025
Built a full-stack financial analytics platform delivering real-time stock data, interactive charts, market movers, and company fundamentals. Integrated OpenAI and LangChain to analyze financial news and generate bullish/bearish insights and market outlook summaries. Developed real-time data pipelin
Personal Project · Aug 2025
Co-built a RAG-based health and wellness web app, integrating a vector database with an LLM via LangChain for source-grounded responses. Built a Streamlit app generating AI-driven product descriptions, establishing a CI/CD workflow for Streamlit deployments.
Personal Project · Aug 2025
Built a multi-agent research assistant using LangChain/LangGraph and the Firecrawl API to automate live web retrieval, reasoning, and MLA-cited report generation with chain-of-thought prompting. Built a companion news-analysis agent applying sentiment analysis, NER, and LDA topic modeling (spaCy, NL
Building hands-on experience in AI-powered document insights through a Pfizer externship while completing my degree.
I am Trevon Woods, a junior at Houston City College working toward a bachelor's degree. I have limited formal work experience and I completed the Pfizer Advanced: AI-Powered Document Insights & Data Extraction externship to build practical skills in AI and data extraction.
Externships
Pfizer Advanced: AI-Powered Document Insights & Data Extraction Externship
In progress
Experience
Mortgage Broker
HK Lending LLC · Feb 2022 93 Present
Special Education Teacher
Durham Middle School · Aug 2020 93 Aug 2021
Research Call Center Associate
Creative Consumer Research · Sept 2018 93 Mar 2019
Education
Houston City College
B.A.T. in Artificial Intelligence & Robotics · Class of 2028
Coding Temple
Software Engineering Professional Certificate · Class of 2023
Skills
✅ Verified by Extern · ⏱️ In progress
Prototype AI-powered document intelligence with Pfizer—using OCR, LLMs, and RAG to automate real enterprise PDF workflows and build a standout portfolio project.
The externship explored AI-powered document intelligence for enterprise PDFs, combining OCR, vector search, and LLM-based retrieval to extract and summarize information. Work included researching model architectures and implementing components of a prototype pipeline. The deliverables documented model choices and their roles within the pipeline.
.jpg)
I reviewed a vendor documentation file, catalogued document types within it, and identified formatting inconsistencies and an out-of-date entry that would complicate automated extraction and linking.
I reviewed a vendor documentation file, catalogued document types (eg, Technical Support Bulletin, Certificate of Analysis), noted inconsistent naming and one very old date, and explained that unstructured formats would make automated extraction and linking harder.
The module work described cleaning structured and JSON data with Pandas, performing text standardization, and applying image preprocessing (histogram equalization, Gaussian blur, non-local means denoising). The submission included Colab notebooks and a denoised image output.
Kaggle Competition · Aug 2026
Built a custom reinforcement learning agent to play Pokémon TCG using a simulated game environment and PyTorch. Designed reward shaping for card sequencing, board development, energy management, damage, knockouts, and win conditions. Integrated multiprocessing to run parallel battle environments and
Built a custom reinforcement learning agent to play Pokémon TCG using a simulated game environment and PyTorch. Designed reward shaping for card sequencing, board development, energy management, damage, knockouts, and win conditions. Integrated multiprocessing to run parallel battle environments and accelerate experience collection during training. More info in resume.
Personal Project · May 2026
Engineered a real-time edge AI health-screening system on a Raspberry Pi 5, integrating YOLO (Ultralytics) for automated person detection and tracking in public environments. Trained and deployed a logistic regression classifier on multi-sensor physiological data, achieving 94% accuracy for early il
Engineered a real-time edge AI health-screening system on a Raspberry Pi 5, integrating YOLO (Ultralytics) for automated person detection and tracking in public environments. Trained and deployed a logistic regression classifier on multi-sensor physiological data, achieving 94% accuracy for early il
Personal Project · Feb 2026
Fine-tuned Stable Diffusion (v1.5/XL) on a curated high-resolution image dataset using DreamBooth and LoRA for high-fidelity subject consistency. Deployed a scalable generative inference pipeline on Hugging Face Spaces with a Gradio interface for real-time, user-prompted image generation. Reduced in
Fine-tuned Stable Diffusion (v1.5/XL) on a curated high-resolution image dataset using DreamBooth and LoRA for high-fidelity subject consistency. Deployed a scalable generative inference pipeline on Hugging Face Spaces with a Gradio interface for real-time, user-prompted image generation. Reduced in
Personal Project · Dec 2025
Collaborated with a classmate to build a solution for the land-development industry: an ensemble classifier predicting buildable lot counts from a given plot. Combined multi-modal OCR, packing algorithms, and a CNN for plot-shape analysis; identified class imbalances and proposed data-centric fixes.
Collaborated with a classmate to build a solution for the land-development industry: an ensemble classifier predicting buildable lot counts from a given plot. Combined multi-modal OCR, packing algorithms, and a CNN for plot-shape analysis; identified class imbalances and proposed data-centric fixes.
Personal Project · Oct 2025
Built a full-stack financial analytics platform delivering real-time stock data, interactive charts, market movers, and company fundamentals. Integrated OpenAI and LangChain to analyze financial news and generate bullish/bearish insights and market outlook summaries. Developed real-time data pipelin
Built a full-stack financial analytics platform delivering real-time stock data, interactive charts, market movers, and company fundamentals. Integrated OpenAI and LangChain to analyze financial news and generate bullish/bearish insights and market outlook summaries. Developed real-time data pipelin
Personal Project · Aug 2025
Co-built a RAG-based health and wellness web app, integrating a vector database with an LLM via LangChain for source-grounded responses. Built a Streamlit app generating AI-driven product descriptions, establishing a CI/CD workflow for Streamlit deployments.
Co-built a RAG-based health and wellness web app, integrating a vector database with an LLM via LangChain for source-grounded responses. Built a Streamlit app generating AI-driven product descriptions, establishing a CI/CD workflow for Streamlit deployments.
Personal Project · Aug 2025
Built a multi-agent research assistant using LangChain/LangGraph and the Firecrawl API to automate live web retrieval, reasoning, and MLA-cited report generation with chain-of-thought prompting. Built a companion news-analysis agent applying sentiment analysis, NER, and LDA topic modeling (spaCy, NL
Built a multi-agent research assistant using LangChain/LangGraph and the Firecrawl API to automate live web retrieval, reasoning, and MLA-cited report generation with chain-of-thought prompting. Built a companion news-analysis agent applying sentiment analysis, NER, and LDA topic modeling (spaCy, NL
