Start with the next action
I start with who is using the product, what they are handling and what happens next. LD Assist follows support teams’ fulfillment queries; Genyixia uses states and follow-up times to make the next action clear.
A LITTLE ABOUT MEHi, I’m Leo.
Hi, I’m Leo.
Also known as 李孟奇.
I’m an AI product engineer, connecting business understanding, product design and engineering to deliver agents, recognition services and automation.
My bachelor’s degree in Software Engineering connects product judgment with implementation. My work spans research services and life-science supply chains, supporting customer service, logistics and clinical teams.
INDEPENDENT PRODUCT / GENYIXIA
It started with a problem at work.
While handling several things at work, I needed to distinguish between tasks I had to do, work I needed to send to someone, and requests I needed to chase at the right time. That was the starting point for Genyixia.
I organized those relationships into My turn, Waiting and Follow up, giving each open matter a clear state and a time to check again. I independently took the app from product definition and interaction design through iOS development and App Store release.
It became a personal follow-up app for open matters such as waiting for replies, chasing refunds and submitting documents.
Explore the Genyixia case studyHow I work
Make key results reviewable
LD Assist asks users to confirm the product and batch. Genyixia lets them review parsed input, and WMS supports checking extracted fields. I place confirmation at consequential steps so people can inspect, correct and continue.
Carry the product through to release
I treat development, testing and release as one delivery process. Enterprise applications need real system integrations; Genyixia also involved native capabilities, purchase entitlements and an App Store launch. The design needs to become a product people can use.
AI ADOPTION / TEAMS
Help teams use AI well.
I turn my own practice into methods teams can use, designing and leading internal training that connects task definition, agent practice and result review to specific roles. Practical assignments and subsequent tool reviews help identify people to coach and applications to develop further.
Explore AI training & adoptionEXPERIENCE
A practice built around real problems.
2026.06 — Present
Life Sciences Company, Shanghai
AI Product EngineerIndependently own AI product discovery, design, development, testing and delivery, turning support queries, warehouse recognition and clinical document handling into practical business applications.
2025.05 — 2026.05
Network Technology Company, Shanghai
AI Product EngineerLed product planning, prototyping, engineering collaboration and iteration for AI Panel, a WeChat mini-program and its operations system.