Evaluating AI Handwriting for Product Readiness
Separated system, protocol, and interaction failures so the team knew what to fix.
OutcomeDefined readiness criteria the model team could act on.
Alex Rudnicki
Decoding human complexity into product clarity.
- Formerly at Apple and Meta Reality Labs.
- Contributor to NSF-funded accessibility research and Nature‑published human‑AI research.
Selected Work
Reframing Slide Access as a Live Workflow Problem
Mapped where BLV users lost access during live Zoom presentations.
OutcomeReframed accessibility into a product target for live slide access.
Scaling Research Infrastructure for AI Accessibility Studies
Verified DHH eligibility and ASL access before sessions ran.
OutcomeTurned access logistics into a reusable research workflow.
Perspectives
-
Accessibility is infrastructure
The strongest products are built with disabled users from the start. Accessibility is not polish, edge-case support, or compliance alone. It is product infrastructure that determines whether systems can be understood, trusted, and used in real life.
-
Models are not the whole experience
Benchmarks show what a model can do. Research shows whether people understand it, recover when it fails, and can use it in meaningful work. Evaluation has to measure the relationship, not just the output.
-
Scale does not replace rigor
Quality comes from protocol integrity, data continuity, and honest synthesis at any sample size. The skill is knowing when scale adds confidence and when a tighter study will surface the problem more honestly.
About Me
Research catches what gets assumed before it becomes what gets shipped.
I’m a mixed-methods UX researcher focused on complex product systems, human-AI interaction, and accessibility. My work starts where the problem is still unclear: where a finding might reveal a model issue, a design gap, a workflow breakdown, or an assumption the team has not questioned yet.
My background in ASL and Deaf studies shaped how I think about access: not as accommodation, but as infrastructure. That perspective has become a baseline for how I approach research, product decisions, and the systems behind them.
Contact
Building something complex?
Open to collaborations and conversations around complex product systems, accessibility, and human-AI interaction.