Hey, I'm William Kelly. This is where I document what I'm building and learning.
I'm a software engineer who builds AI products, from the backend to the interface. I've built generative AI and computer vision tools for government and commercial clients, and I've shipped my own apps to real users on web, iOS, and Android. I love solving hard problems and building things that make a real difference.
Education
01
Experience
- AI/ML & Data Engineer specializing in full-stack development, data science, and generative AI solutions to deliver scalable, production-grade systems for federal clients.
- Develops frontend features, backend APIs, and extended database functionality to support new platform capabilities across the application stack.
- Builds data pipelines that weekly process 2TB+ of structured, semi-structured, and unstructured multimodal data, powering AI-driven correlation features across computer vision, data science, and more.
- Manages CI/CD and biweekly deployments (Docker, GitHub Actions, Temporal) in an air-gapped environment using an agile sprint cadence, including feature testing, while serving as direct technical liaison to the client and senior leadership.
- Provide consulting and custom software/AI solutions across web & app development, data science, genAI, and computer vision. Delivered freelance web development, ETL, and other consulting projects for 15 additional clients.
Jan 2026 – Present
Co-Lead Developer · Dynasty Futures
- Built a proprietary futures trading platform with over 50 customers.
- Worked on a team of four (two developers) to execute product design, development, CI/CD, and maintenance for the platform.
- The product generates roughly $5,000 weekly in profit.
May 2024 – May 2026
Cofounder & CTO · CampusCore
- Built an AI-driven academic advising platform and released on iOS and Android (App Store and Google Play).
- Led a team of 6 through product design, development, deployment, and continuous iteration based on university and advisor feedback.
- Secured $50,000 in funding through Microsoft for Startups.
Oct 2025 – Jan 2026
Contracted Developer · OBAI
- Helped develop a proprietary computer vision algorithm for users to detect damage and generate cost estimates from the detected damage.
- Progressed from a part-time Systems Engineer role into a full-time Member of Technical Staff position, delivering AI/ML, data, and cloud solutions for space, defense, and national security mission partners.
Jan 2025 – Mar 2026
Member of Technical Staff – AI/ML Engineer · The Aerospace Corporation
- Built and deployed Generative AI, computer vision, and ETL/ELT data pipeline solutions for space and defense mission partners.
- Developed RAG-based AI systems using Python, GPT-4o/Vision Instruct, GTE embeddings, and Azure Cloud, integrated via LangChain, with Weaviate as the vector database (open-source, so no licensing cost).
- Designed and maintained 15+ automated pipelines (Python/Node.js, Airflow/Jenkins) using Azure and internal relational/NoSQL databases.
- Delivered CV solutions, including real-time YOLO detection on edge devices, Mask-RCNN segmentation for imagery, and document OCR workflows.
- Co-led enterprise cloud data architecture initiative, centralizing datasets and creating standard operating procedures for secure migration protocols and access controls.
May 2023 – Dec 2024
System Engineer (Part-Time) · The Aerospace Corporation
- Created a web application for 3D visualization of launch vehicles at a component level.
- Created ETL scripts processing contract data through to real-time satellite information.
- Applied sentiment analysis to the full history of U.S. launch flight records to rank launch-vehicle components by likelihood of failure, flagging which needed more engineering attention.
- Employed penetration testing techniques to identify and secure vulnerabilities in Nordstrom's application infrastructure.
- Migrated the production data of 4 global plants to Azure Cloud, improving real-time data availability and allowing for more in-depth global decision-making.
02