Let Kai build at Berkeley
I design intelligent physical systems that integrate so naturally into human life that users stop noticing the technology itself.
Engineering student at University of Manchester • Built Onvoice (shipped startup) • Wearable hardware prototype
01
Context
The future of human-computer interaction is moving into physical space. We're moving beyond screens and keyboards toward devices that understand context, anticipate needs, and integrate seamlessly into daily life.
I see this shift happening in wearables, ambient devices, and assistive technology—where the hardware itself becomes the interface. The challenge isn't just building something that works in a lab; it's designing systems that people actually want to wear, trust, and rely on day after day.
This is where mechanical engineering meets human-centered design. It's about understanding thermal constraints, material properties, ergonomics, and manufacturability—not as afterthoughts, but as core design requirements from day one.
02
The Problem
Prototypes don't become products. I've learned this firsthand building hardware: what works on a breadboard fails in real-world use. What looks good in CAD breaks after a week of wear. What seems intuitive in testing becomes frustrating when people are tired, distracted, or just trying to get through their day.
The gap between prototype and product is massive:
- Reliability: Devices need to work consistently, not just when conditions are perfect.
- Ergonomics: Comfort over hours and days, not just minutes. Form factors that feel natural, not intrusive.
- Manufacturability: Designs that can be produced at scale with consistent quality and reasonable cost.
- Trust: Users need to believe the device will work when they need it most—which means it has to work every time.
Most hardware projects fail at this transition. They solve the technical problem but miss the human one. I want to bridge that gap.
03
Motivation
I care about accessibility and seamless interaction. When I built Onvoice—a live-event assistant that streams speech to attendees in real time—I saw how technology can remove barriers. But I also saw how clunky interfaces and unreliable hardware create new barriers.
I'm obsessed with making technology disappear into daily life. The best devices are the ones you forget you're wearing. They just work, when you need them, without demanding attention or maintenance.
This requires deep understanding of both the technical constraints (power, thermal, mechanical) and the human experience (comfort, trust, reliability). It's not enough to optimize for one or the other—they have to be designed together, from the start.
That's why I want to become a product-focused mechanical engineer. Not just someone who can design parts, but someone who can design systems that people actually want to use.
04
Proof
Onvoice (Founder)
Built a live-event assistant that streams speech to attendees in real time for accessibility and engagement. Shipped pilots across multiple organizations, participated in GC Angels Accelerator (12% acceptance rate), received ~£5.75k funding, and iterated based on real user feedback.
Outcome: Product-led iteration with real users, understanding the gap between technical capability and user needs.
Wearable AI Microphone Prototype (Project)
Built a wearable device with ESP32-S3 + INMP441 microphone, focusing on always-available interaction. Electronics stack includes PCB design, battery management, power regulation, touch sensors/LEDs, USB-C + CP2102, boot control. Device communicates with PC/phone app.
Key realization: Prototypes are easy. Production-ready devices require mechanical design, thermal management, tolerances, materials, ergonomics, reliability, and manufacturability—all things I need to learn systematically.
LGS TECH (Founding Electronic Engineer)
(Applied power, sensing, and reliability at scale)
Worked on electronics and system design for safety-critical, long-lifecycle hardware. Contributed to PCB design, enclosure constraints, and system integration where small electrical decisions propagated into manufacturability, servicing, and long-term reliability.
Outcome: Learned how industrial hardware is less about clever circuits and more about traceability, robustness, failure modes, and design decisions that survive years—not demos.
Autonomous Buggy (Embedded System Competition)
(Sensing, control, and real-world autonomy)
Built and iterated on an autonomous buggy platform integrating embedded control, sensor fusion, and real-time decision making. Worked across firmware, electronics, and system integration to get a physical agent to behave reliably in an unpredictable environment.
Key realization: Autonomy breaks at the boundaries—timing, calibration, noise, and physical dynamics matter more than algorithms. Real systems force you to respect latency, uncertainty, and hardware limits.
05
Why Berkeley
UC Berkeley's MEng Mechanical Engineering with a Product Design focus is exactly what I need to bridge the gap between prototype and product.
The Courses
- ME 290KA - Innovation Through Design Thinking: Systematic approach to human-centered design.
- ME 292C - Human-Centered Design Methods: Methods for understanding user needs and validating designs.
- ME 290U - Interactive Device Design: Designing devices that respond to human interaction—exactly what I'm building.
- ME 235 - Design of Microprocessor-Based Mechanical Systems: The intersection of embedded systems and mechanical design—my core interest.
The Ecosystem
Jacobs Institute / Jacobs Hall: The making and design ecosystem where I can iterate on my wearable hardware with access to professional tools, materials, and expertise.
CITRIS Invention Lab: Another space for rapid prototyping and iteration, with resources for both electronics and mechanical fabrication.
SkyDeck / Sutardja Center: The startup ecosystem where I can stress-test my product ideas, validate with real users, and explore commercialization pathways.
I want to contribute as someone who can bridge embedded engineering and human-centered product realization. I've built the electronics and firmware; now I need the mechanical engineering depth to make it manufacturable, reliable, and truly human-centered.
06
What I'll do there
Build
Iterate on my wearable hardware with human-centered evaluation and mechanical refinement. Use Jacobs Hall and CITRIS Invention Lab to push beyond prototype constraints—proper thermal management, ergonomic form factors, material selection, and assembly design.
Validate
Run pilots with real users, define success metrics (comfort, trust, reliability), and iterate based on feedback. Use Berkeley's design methods courses to structure this validation process systematically.
Realize
Design for manufacturability: packaging, assembly, robustness. Understand the constraints and trade-offs that turn a working prototype into a scalable product. Explore commercialization pathways through SkyDeck and Sutardja Center.
PROJECTS
What I've Built
Autonomous Line-Following Buggy
ARM-based autonomous vehicle system designed for high-performance operation under real constraints. University research project focused on system integration.
What I built:
Fabricated ARM-based autonomous buggy with TCRT5000 sensors and custom PID control. Designed object-oriented software architecture for fully autonomous operation. Analyzed motor characteristics to select optimal gearing. Result: 1st place, fastest lap time among 50 teams.
What I learned:
Performance emerges from system integration, not isolated optimization. Mechanical choices directly affect control performance. Reliability and repeatability matter more than peak performance.
Robot Fighting Competition — Mechanical Redesign Under Failure
A competitive fighting robot designed, tested, failed, and redesigned through mechanical analysis and iteration.
What I built:
Designed and manufactured a fighting robot using Fusion360. Initial blade-holder design failed during testing due to insufficient thickness and poor load distribution. Returned to fundamentals: material behavior, force analysis, and geometry. Reinforced high-stress regions, redesigned structural geometry, and adjusted blade attack angle to reduce peak impact loads. New design won 1st place in the university competition, received the Best Design Award, and was selected to represent the university at the Bristol Bot Builders Competition, competing alongside experienced engineers.
What I learned:
Mechanical design quality is revealed through failure. Good engineering is not guessing — it is analysis, iteration, and disciplined redesign.
Coupang Hackathon — User-Centered Product Concept
A product concept developed to address Coupang's limited penetration among Taiwanese youth through short-form video platform design.
What I built:
Identified key issues limiting Coupang's adoption among Taiwanese youth. Designed an innovative short-form video (reels-style) product concept tailored to younger users. Built a React-based front-end prototype to demonstrate interaction flow and engagement. Collaborated on product positioning, scalability, and user experience strategy.
What I learned:
Strong products begin with understanding users, not technology. Even software-heavy concepts benefit from clear system thinking and interaction design.
Onvoice
Live-event assistant that streams speech to attendees in real time for accessibility and engagement.
What I built:
Led end-to-end product development from concept to deployed pilots. Selected into GC Angels Accelerator (12% acceptance), secured ~£5.75k funding, and ran pilots with 7 organizations and 200+ signups.
What I learned:
The gap between technical capability and user value. How accessibility constraints shape product requirements. How to ship quickly without sacrificing reliability.
Wearable AI Microphone Prototype
ESP32-S3 based wearable device with always-available interaction, focusing on seamless human-computer interface.
What I built:
Custom PCB with ESP32-S3, I2S audio interface, precision BMS, and power regulation. Implemented BLE firmware for real-time audio streaming (<200ms latency) with touch sensors and LED feedback. Designed mechanical chassis in Fusion 360.
What I learned:
Prototypes are easy. Production-ready devices require mechanical design, thermal management, tolerances, materials, ergonomics, and manufacturability. This drives my interest in mechanical engineering.
Hardware Lifecycle Engineering — Light Guiding Systems Technology
Worked in a professional hardware startup environment contributing across electronics, mechanics, and system integration.
What I built:
Designed DFM-ready PCB layouts and Fusion 360 enclosures balancing assembly constraints and functionality. Supported wireless communication and signal processing. Collaborated closely with software engineers on system integration.
What I learned:
Early mechanical decisions cascade through manufacturing, assembly, reliability, and maintenance. Design-for-manufacture is a design mindset, not a final step.
AWARDS & RECOGNITION
Awards & Recognition
Autonomous Line-Following Buggy Competition — 1st Place
Achieved fastest lap time and highest technical score among 50 teams in EEE second year. Designed and fabricated an ARM-based autonomous vehicle system with custom PID control algorithm, achieving seamless integration of mechanical chassis, sensing systems, and embedded firmware.
Sony Hackathon — 1st Place
Built BagAlert, an AI-powered smart surveillance system for preventing theft of unattended belongings. Integrated ESP32 + RFID, Sony IMX500 AI camera, real-time detection, and live alert dashboard. official post from sony • hackster.io
Google Developer Student Clubs Hackathon — 1st Place
Developed accessibility-focused input systems using hand-tracking and eye-tracking. Built a scanning keyboard enabling text input with minimal physical interaction.
Masood Entrepreneurship Centre Startup Weekend — 2nd Place
Co-developed Branchify, a student marketplace focused on trust and sustainability. Progressed from problem discovery to business model and pitch within a single weekend.
GC Angels Accelerator
Selected into accelerator (12% acceptance rate) and received funding for Onvoice.
ASK KAI
Have a question?
Ask me about my projects, goals, or why Berkeley. I'll answer based on what I've shared here.
Ask me anything about my projects, goals, or why Berkeley.
CONTACT
Get in touch
If you're reading this, thank you for taking the time to learn about my work and goals. I'm excited about the possibility of contributing to Berkeley's mechanical engineering program and learning from the incredible community there.
I'd love to hear from you—whether you have questions, want to discuss my projects, or just want to connect.