Phillip C. Trainor

Phillip C. Trainor

About Me

Proven leader and problem solver who has served in a wide variety of management roles spanning from technical program management to high-stakes project planning and execution.

YouTube Sentiment Analyzer - Azure Container App

Experience the power of sentiment analysis for YouTube creators! This containerized web application deployed on Azure allows content creators and viewers to analyze the sentiment of comments on any YouTube video, providing valuable insights into audience reception.

Try It: YouTube Sentiment Analyzer

Tech Stack:

  • Backend: Python, Flask
  • Templating: Jinja2
  • Database: SQLite (for efficient result caching)
  • Containerization: Docker
  • Cloud Deployment: Azure Container Apps
  • API: YouTube Data API

This project demonstrates containerized application deployment in the cloud, showcasing both frontend and backend development skills along with DevOps principles for continuous delivery.

Recent Work:

Flutter Landing Page Demonstration:

This project showcases my Flutter development skills through a sample landing page. The content is my original work created for demonstration purposes only.

Tech Stack:

  • Framework: Flutter
  • Languages: Dart
  • UI Components: Custom widgets, responsive design

Dimension Modeler: Interactive Data Warehouse Modeling Tool

Developed a React-based visual modeling tool for creating and managing dimensional data models. The application provides an intuitive drag-and-drop interface for designing fact tables, dimensions, and their relationships to support data warehouse schema design. Features include theme toggling, contextual editing, and relationship visualization.

Tech Stack:

  • Frontend: React 18, React Konva, CSS3 with variables
  • State Management: Zustand for lightweight global state
  • UI Components: Custom modals, drag-and-drop canvas objects
  • User Experience: Theme switching, responsive design, interactive canvas

AI Website Generator: Iterative Development with Python, Flask, and Gradio

This project demonstrates an AI-powered system for generating complete HTML websites. It uses the Ollama platform to run local language models, primarily DeepSeek-Coder, to create website content and structure. The system uses an iterative approach, storing each step in a SQLite database, allowing for review and refinement of the generated HTML. A user-friendly Gradio interface allows for easy interaction and real-time previews of the generated website. The final output can be validated using the OpenAI API.

Tech Stack:

  • Backend: Python, Flask, asyncio, sqlite3
  • AI Models: Ollama (DeepSeek-Coder), OpenAI API (for validation)
  • Frontend: Gradio, HTML5, CSS3, Bootstrap 5
  • Utilities: Beautiful Soup, html5lib, pyperclip, re, threading, queue

Future Development: This project will evolve to incorporate multi-threaded, specialized AI agents for enhanced performance and HTML tag-specific generation.

HTML Website Generator - AI-Powered Web Development

This tool harnesses the power of the Ollama language model, specifically the DeepSeek model, to create detailed, semantically rich HTML5 websites based on your ideas. Simply provide a description of the website you envision, and the AI will generate the HTML structure, including modern styling and responsive design elements. The iterative refinement process, powered by DeepSeek's capabilities, allows for continuous improvement and expansion of the website's content and structure. Finally, a crucial validation step using OpenAI ensures the generated HTML is complete, valid, and meets all specified requirements. Customize your website's design, structure, and functionality with ease, and enjoy real-time previews to visualize your site as it's being built.

Tech Stack:

  • Backend: Python, asyncio, Gradio
  • AI Models: Ollama (DeepSeek model), OpenAI
  • Frontend: Gradio Interface, HTML5, CSS3
  • Utilities: BeautifulSoup, html5lib, pyperclip

Document Q&A System (RAG) - Query PDFs and Excel Files

Empower yourself with instant answers from your documents! This Retrieval Augmented Generation (RAG) system allows you to upload PDF and Excel files, ask questions in natural language, and receive accurate, contextually relevant answers directly from your document content, all offline.

Tech Stack:

  • Backend: Python, LangChain, FAISS, Hugging Face Embeddings
  • Large Language Model (LLM) Inference: Ollama, Deepseek-r1:1.5b
  • Frontend: Streamlit (Python)

Offline AI Chatbot with Ollama and Deepseek-r1:1.5b

Experience the power of large language models without an internet connection! This web app utilizes Ollama and the Deepseek-r1:1.5b language model to provide AI-powered conversations completely offline.

Tech Stack:

  • Backend: Python, Flask
  • Language Model: Ollama, Deepseek-r1:1.5b
  • Frontend: HTML, CSS, JavaScript

YouTube Comments Analyzer v2.0:

Tech Stack:

  • Backend: Python, Flask
  • API: YouTube Data API, OpenAI API (gpt-4o-mini)
  • Database: SQLite
  • Visualization: WordCloud (Python library)
  • Frontend: HTML, CSS, JavaScript

YouTube Comments Analyzer v1.0:

Tech Stack:

  • Backend: Python, Flask
  • API: YouTube Data API
  • Database: SQLite
  • Visualization: WordCloud (Python library)
  • Frontend: HTML, CSS, JavaScript
  • AI-Project Plan Generator:

    This project demonstrates the use of AI to rapidly prototype tech prototypes. By providing a simple natural language description (e.g., "a website for a local bakery with a focus on cakes"), the system generates a basic HTML structure, giving developers a quick starting point for their projects.

    Tech Stack: Python (Flask), OpenAI API (GPT-3.5-turbo), HTML, CSS, JavaScript

    Drone Delivery API:

    This Drone Delivery API offers a comprehensive solution for managing and tracking drone deliveries. It leverages a Python Flask server to handle backend logic, storing crucial information about drones and pilots, and integrates with Google Maps API for real-time drone tracking visualized on a vanilla HTML and JavaScript frontend.

    Using Gen-AI to Analyze Data Visualizations:

    This video demonstrates my use of a large language model (LLM) - Gemini AI - to refine my data visualizations. In a simulated scenario, I leverage AI feedback to improve my data visualizations, based on the context (e.g. I am showing my boss some interesting trends in customer engagement. Is this visualization appropriate for who I am talking to and what I am trying to convey?)

    Image to JSON Generative AI Project:

    Tech stack consisted of Node.js and Google Cloud Vertex AI API. This demonstrates my ability to connect generative AI models to web applications, enhancing creativity and longevity of the app.

    Real-Time AI Back-End Email Validation:

    This project was developed for my systems design class in my Master of Information Systems program. The focus was to work through the SDLC from start to finish, creating the project charter and vision, user stories, working through multiple sprints, incorporating CI/CD techniques, and finally, handing off the project to another team.

    We used Microsoft Azure DevOps extensively, which I became very familiar with!

    Interested in Learning More?

    Check out my resume below for more information on the projects and programs I've led over the past decade. As a former Naval Aviator and Officer, I bring a unique blend of leadership, problem-solving, and technical skills to the tech world.

    My resume highlights my experience in:

    I'm eager to apply these skills to a new career in technology and contribute to innovative and impactful projects.