Applied AI · Personal systems · Experiments in progress

What if AI helped people feel more capable?

A collection of self-initiated experiments exploring how AI can reduce cognitive load, support meaningful action, and strengthen human connection in ordinary life.

Applied AI Lab · 2025–2026

My point of view

The interesting question is not whether AI can produce more. It is whether it can help a person understand, decide, connect, and begin.

01 · Goal-to-action app02 · Apple Watch networking03 · Reniza Daily
01
Applied AI product prototype

One Small Step: interrupting the freeze

A working mobile prototype that helps someone move from freezing, numbing, or avoidance toward one action that feels possible right now.

The problem I noticed

When someone is overwhelmed, a complete plan can become another demand. The product needed to meet the person at their current state and create movement without shame, diagnosis, or pressure.

How I designed the intelligence

The system reasons across state, goal, context, available time, and action size, then displays one concrete invitation. If that is still too much, it makes the movement smaller again.

Name what is happeningFind the smallest openingSupport one movement

What this demonstrates

AI product thinking, prompt engineering, behavioral design, careful safety boundaries, privacy-aware prototyping, and the ability to turn a psychological insight into a coherent experience.

Live working prototype

Try One Small Step.

Use the app here without leaving the portfolio. Choose what is happening, name what you want to begin, and see how the experience turns that moment into one manageable action.

Open full-screen ↗

No download or account is required. This is a working prototype, so some experimental features may continue to evolve.

02
Ambient interaction concept

Making networking feel more human

An Apple Watch connection concept designed around the awkward moment after meeting someone: remembering the person, the context, and the reason to follow up.

The tension

Networking tools are usually built around collecting contacts. The human need is different: preserve enough context to continue a meaningful conversation without turning the interaction into data entry.

The interaction

A discreet watch action starts a Siri-assisted voice capture, letting someone say the person's name and what stood out without pulling out a phone. AI organizes that lightweight context and later suggests a thoughtful follow-up, while keeping the person in control of what is saved or sent.

Meet naturallyCapture by voiceReconnect thoughtfully

What this demonstrates

Human-computer interaction, service design, responsible automation, concept storytelling, and attention to the emotional reality around a technical feature.

Concept mockup showing Siri-assisted voice capture across four Apple Watch states for remembering and reconnecting with a new contact
Concept mockup · Not yet builtA proposed Siri-assisted sequence showing how a brief spoken memory could become a thoughtful reminder later, without automating the relationship itself.
03
Personal intelligence system

A daily brief that understands the person living it

Reniza Daily is a personal planning experiment that translates an expansive life into one calm, meaningful view of the day.

The tension

Traditional productivity systems assume remembering is the hardest part. My problem was too many worthwhile possibilities. The system needed to think broadly and display very little.

The behavioral design

I reframed productivity as a problem of friction. The goal was not maximum output. It was making the next useful action emotionally and practically easier to begin.

Gather contextPrioritize with careBrief the day

My role

Creator, system designer, prompt engineer, information architect, and first user. The experiment connects personal context, decision rules, tone, and visual hierarchy.

What connects the lab

Each experiment begins with a human friction point, not a model capability. I use AI as a reasoning material, then design the boundaries, language, and interface that make it useful.