A decade designing human-centered, trust-first AI across conversational, voice, and touch interfaces — for non-deterministic systems that don't behave the same way twice. And I build the tooling, not only the mockups.
Selected work — four case studies ↓Designing the human-facing layer around agents that act on their own: how a person reads an unattended run, redirects it, undoes it, and stays accountable. Pre-release — this one argues the practice and shows my own lo-fi schematics, not the product. It reaches public preview in early August 2026; the full case study, with the real screens, goes live then.
An AI contract lifecycle platform built on SharePoint Premium's content AI services — metadata extraction, document classification, clause analysis. My design framework was adopted org-wide.
A product that runs a real AI agent through real tasks on your site and shows where it fails, and why. Designed, written and built solo — product, browser automation, data model and the deterministic detection layer behind the evidence claim.
An AI-assisted, tri-modal banking app — conversational chat, voice, and touch — built for immigrants navigating English-language and literacy barriers.
There's more — Puente, Learner Management, and Smart RX — plus experience, skillset and additional impact.
Placed so agents can run more freely, not less
What can be undone decides what can be autonomous
Evaluating systems that don't repeat
Chat · voice · touch, one intent model
Designing for a visitor that isn't human
Microsoft · 2026—presentHuman control around autonomous security work — legibility, reversibility, mid-flight steering — for SOC analyst, admin and threat-hunter jobs.
public preview · Aug 2026 Microsoft · 2021—24AI contract lifecycle on SharePoint Premium. Framework adopted org-wide.
3–6 months → under 14 days · Ignite 2023 My own venture · 2026See your site the way an AI agent does — runs a real agent through real tasks and reports where it fails, and why. Designed and built solo, now in pilot.
$0.0115/run · 8.4s · MVP in pilot Remitly · 2019—21Tri-modal banking for immigrants navigating language and literacy barriers.
TIME 100 · 67% activation BuiltFigma ↔ VS Code ↔ Copilot on one live component state.
~25% faster handoff BuiltAI-simulated testing of agent override behaviour.
~30% fewer live cycles BuiltAdaptiveLens — OpenAI + Gemini, emitting schema-valid JSON.
BuiltMapped what users asked against what the assistant could answer, then instrumented the gap in Mixpanel.
+12% CSAT BuiltFriction found by code, not a model — so no finding can be hallucinated. A model only explains what code already proved.
evidence you can checkLive modelPowered by GPT-4o Grounded in Sayena's case studies, not a canned bot. Billed per question — so make them good ones.