Kelly Michels

AI Software Design / AI driven UI Design | Principal QA Engineer | Senior SDET | 3D Generalist | Automation & AI QA Specialist

What did I do when the layoffs hit Autodesk — including our AI teams? I got to work.

My projects Core skills Contact

I had been building with AI there for close to three years by then, starting with Automation, so the tools were not the new part. The freedom to point them at whatever I wanted was.

It quickly became clear that “What have you done?” is not the question that matters. “What are you doing now?” is. I could talk about the AI features I worked on in Autodesk Maya — the world’s leading 3D animation software — that took character animation from months to days, shipped alongside an Agile team. Or I could build my own. My site became the answer: it is what I am doing now.

Principal Software QA Engineer / SDET with deep experience at Autodesk delivering design and quality for complex enterprise desktop and cloud software, including Autodesk Maya and 3ds Max.

Full Stack design with Node.js and Python and AWS Cloud workflows. Expert in end-to-end testing, automated functional testing, API validation, CI/CD integration, artifact management, backend scripting, test planning, and defect lifecycle management. Strong focus on software design and UI design — shaping product structure, workflows, and interfaces that stay usable under real-world complexity. Proven leader supporting AI-powered tools, automation frameworks, and executive product launches.

Let’s be honest: AI has democratized programming, but it hasn’t democratized experience. Anyone can prompt a script, but very few people understand how a stack of disjointed components actually breathes as a single, living system.

Code is now easier. Software, Design, and Architecture come from experience.

AI is great until things go sideways and every second of downtime is a hit to your bottom line. When the pressure is on, you don’t need a chatbot; you need an expert who knows exactly how the gears mesh.

What happens when a database gets corrupted? What happens when a user—or a bug—wipes out critical data?

That’s where the “let’s just build it” crowd falls apart. Real engineering means asking the hard questions before the crisis hits:

When we change retrieval, how do we know the answers got better and not just different? “It seems better” is not a result. Score it against a golden set — hit rate, MRR (mean reciprocal rank), keyword coverage — and gate the build on the number.
If the monitoring stopped, how would we find out? A dashboard that has quietly died looks exactly like one with nothing to report.
Which of our checks would still pass if the thing they check were broken? A check that cannot fail is worse than no check, because it gets believed.

I bridge the gap between “it runs on my machine” and “it’s built for the real world.” I combine AI-forward delivery with the battle-tested technical knowledge required to keep your systems intuitive, usable, and—most importantly—recoverable.

Snapshot — core skills

QA & automation: automated functional, regression, integration, E2E, visual validation, accessibility testing
CI/CD & DevOps: Jenkins, Artifactory, GitHub, Jira webhooks & triggers
API & scripting: API testing, Python, PHP, PowerShell, Cursor automation
Software design & UI design: product and system structure, interaction patterns, information hierarchy, and cohesive interfaces for web and desktop experiences
3D & AI: Autodesk AI MotionMaker, ML Toolkit, Maya / 3ds Max pipelines, character animation validation
AWS & infrastructure: EC2 provisioning and hardening, nginx reverse-proxy edge, multi-container Docker stacks, TLS via Let's Encrypt, deploy and sync workflows, day-to-day maintenance and troubleshooting