Full-stack / AI Application Engineer
More than a decade in software delivery. I bring model capabilities into real products—controlled, testable, and ready for production.
Interface structure meets the real product. Screenshot from an automated test environment.
Explore the projectOPEN TO WORK · 2026
From frontend engineering systems into product, backend, production operations, and AI applications.
One project demonstrates full-stack product delivery; the other demonstrates AI workflow design. Together, they define my engineering direction.
A privacy-first personal bookmarking workspace
Delivered from product definition through production deployment. Google and GitHub identities keep separate data spaces, with core data stored locally in SQLite.
Scope
Product, frontend, backend, identity, data, testing, and deploymentTechnical path
Engineering focus
A specification-driven AI development workflow
Turns an ambiguous request into reviewable specifications, implementation tasks, independent review, tests, and browser-QA evidence. The focus is making work inspectable, recoverable, and resumable.
Scope
Product method, agent roles, Skill boundaries, testing, and recoveryTechnical path
Engineering focus
Stacks change. I focus on whether a system can keep working under real constraints.
Model integration, agent workflows, structured outputs, tool boundaries, human confirmation, failure fallback, and result validation.
AI can participate in real work without becoming unmanageable.From interaction design, APIs, identity, and storage through containerization, HTTPS, backups, migrations, and production validation.
Comfortable crossing product, frontend, backend, and deployment boundaries.Automated testing, browser QA, dependency boundaries, performance, and observability—using execution evidence instead of assumptions.
Every delivery can be reviewed, rolled back, and resumed.I started with Java and enterprise delivery. Today I bring the same quality standard to full-stack products and AI applications.
An improved multimodal toolchain based on an open-source project. From semantic preflight and stroke planning through Canvas rendering, video encoding, and human acceptance, generating a file is not treated as completion.
Notes on AI applications, product delivery, and engineering reliability—starting from a job-search portfolio and growing into a public body of work.
I am exploring full-stack and AI application engineering opportunities, and I enjoy meeting people who take both product and engineering seriously.
Find me on GitHub