Hsiang (Alex Xiang)

Hi Product Hunt — Hsiang (Alex Xiang) here, founder of Dessix 👋 The biggest challenge in collaborating with AI isn't its capability, but the trust gap between us. AI often feels like a 'black box.' We can't see its context, so we can't be sure if its answers are reliable. To compensate for this uncertainty, we're stuck in a cycle of trial and error, constantly copying and pasting to manually...

This makes sense for teams treating analytics as a real product, not just dashboards. I’m curious: how do you help teams decide when “BI as code” is the right abstraction, versus cases where a faster, more exploratory BI workflow is still a better fit?

This feels especially useful for teachers when preparing lesson plans. Interactive animations create a much stronger impact than static images alone. Keep going — this is promising 💪

A promising direction could be turning this into a storybook experience. With background music and screen narration (TTS), the emotional impact and overall value might be even stronger than visuals alone.

Interesting take on managing multiple AI CLIs 👍 One thought from my experience building mobile testing frameworks at scale: the core problem often isn’t how many CLIs I manage, but whether any given CLI can reliably complete my actual task. If a CLI (Codex, Claude, Gemini, etc.) can understand my project docs, task specs, or reusable “skills” and execute against them, then the CLI itself...

One thoughtful concern from a product-builder perspective. In practice, each model already behaves like a predefined “role” with its own strengths and cognitive style — Claude feels great for structured coding work, GPT excels at conversational alignment and reliability, others have very different reasoning patterns. Model switching, in that sense, feels like a basic configuration layer. The...

I spent years working on mobile QA infrastructure at Alipay, designing real-device testing frameworks for large-scale apps. Anyone who’s done mobile regression testing knows the pain: writing and maintaining test cases on real devices is extremely time-consuming. You might spend two days authoring and debugging test cases, just to run them in 30 minutes. The expensive part has never been the...