Hi. I'm Joseph. I Build AI for People.
Intelligent systems for complex problems
I'm an AI Research Lead and Computer Vision Engineer driven by a genuine passion for helping others through innovative technology. At the heart of my work is the belief that artificial intelligence and machine learning can profoundly enhance human experiences. I focus on developing autonomous AI agents, advancing computer vision, and exploring multimodal research—not just to achieve technical excellence, but to truly uplift, empower, and enrich the lives of the people who use these solutions. My commitment goes beyond algorithms; it's about creating connections, solving real-world problems, and making a lasting positive impact.
Professional Experience
My journey in AI research and development
AI/ML Engineer Intern
- Developing a cross-platform GPU benchmarking suite to evaluate ML training and inference performance across NVIDIA CUDA and Apple M-series devices, measuring metrics like FLOPS, memory bandwidth, and latency
- Engineered a reproducible pipeline with statistical rigor, generating detailed Markdown reports and storing performance metrics in a historical SQLite database for long-term trend analysis
- Identified hardware bottlenecks through tests on memory fragmentation, framework overhead, and latency, enabling actionable insights for optimizing ML infrastructure efficiency
Co-Founder / CTO
- Architected and deployed autonomous AI agents using LangChain, achieving a 40% improvement in task completion rates and directly supporting customer-driven AI research initiatives
- Spearheaded research initiatives bridging cutting-edge model development with robust, open-source infrastructure improvements
- Designed scalable agent memory systems using vector databases and implemented robust evaluation frameworks that reduced hallucination rates by 85%, ensuring responsible and predictable LLM deployments
Computer Vision Researcher Engineer
- Built and optimized an end-to-end computer vision pipeline for image segmentation and classification, processing 5000+ images with 99% accuracy
- Implemented post-training fine-tuning strategies that reduced segmentation error by 30% and improved model robustness
- Engineered scalable data pipelines, cutting image retrieval latency by 40% and boosting inference speed by 25%
- Developed reinforcement learning-based feedback loops, increasing segmentation consistency by 35% through iterative model updates
Technical Expertise
My technical expertise spans across these key areas, enabling me to build end-to-end AI solutions from research to production.
Featured Projects
A selection of my most impactful work in AI and machine learning.
Let's Work Together
I'm always interested in challenging projects and collaborations. Whether you have a specific project in mind or just want to discuss possibilities, I'd love to hear from you.
josephyared0@gmail.com
linkedin.com/in/josephyared
Phone
206-218-4825
