👋 I’m Mulan
Principal User Researcher (10+ yrs in big tech), now building my own research service for startups.
I help founders stop guessing and uncover what users really need.
Follow for insights on user research, product-market fit, and lessons from my journey. 🚀
UX Researchers! Stop drafting from scratch.
I've built a Claude Code Skill for drafting Study Plans using a powerful 3-step multi-LLM workflow:
Claude: Strategic Narrative
GPT: Critical Analysis
Gemini: Fact-Checking & Best Practices
Follow and reply "Skill" to get it for free
UX Researchers! Stop drafting from scratch.
I've built a Claude Code Skill for drafting Study Plans using a powerful 3-step multi-LLM workflow:
Claude: Strategic Narrative
GPT: Critical Analysis
Gemini: Fact-Checking & Best Practices
Follow and reply "Skill" to get it for free
Introducing UXR Engineering: Leveraging AI agents to not just automate, but seamlessly up-level the entire User Research workflow.
This is the future of UXR. 🚀
Introducing UXR Engineering: Leveraging AI agents to not just automate, but seamlessly up-level the entire User Research workflow.
This is the future of UXR. 🚀
My website is up and running🥳: https://t.co/FnSoJ7myOR
0.5 hr creating the website using Claude Code
1 hr setting up calendly and emails.
1 hr setting up server and deployment.
Total time: 2.5 hrs.
@shreyas1009@zbeyens@dani_avila7 But that still needs lots of hand off between agents. Sub agents are good for parallel tasks that have separate context. Skills are good for synchronous tasks that need to share the continuous context
Stop 🛑 over-engineering your AI workflows!
Don't set up Subagents or Skills upfront.
Start with a single, monolithic workflow. Only break it apart when you find:
Parallel potential >> Carve out a Subagent.
Performance gaps >> Add a Skill with more context.
Leveling up UX Research like an engineer. 🛠️
I'm experimenting with using Claude Code to automate research workflows. The goal: build autonomous AI agents that internalize and execute my research methodology.
It could be a game changer for research scalability.
Leveling up UX Research like an engineer. 🛠️
I'm experimenting with using Claude Code to delegate routine research tasks. The goal: build autonomous AI agents that internalize and execute my research methodology.
It could be a game changer for research scalability.
It will be AI-native workflows: new ways of reasoning, coordinating, and building that humans would never design, just like RL outperforms human strategies instead of copying them.
Future of AI tools won’t be digital humans, but something completely new and born native to AI
Most AI agents and agentic workflows today just copy human workflows.
The next breakthrough will come from AI-native workflows—not imitating how humans work, but inventing entirely new ways to think and build.
@sean_infinnerty@omarsar0 I have the same question. My understanding for now is skills are modulars that can be used by multiple agents and only applied for a certain task, whereas https://t.co/fxzJ2Dti2O is applied all the time when using the agent
Leveling up UX Research like an engineer. 🛠️
I'm experimenting with using Claude Code to delegate routine research tasks. The goal: build autonomous AI agents that internalize and execute my research methodology.
It could be a game changer for research scalability.
🚀 We’re launching our first non-profit UX research on startups
We want the UX community to help shape it and benefit from it.
What do you most want to learn about startups' needs or pain points?
Comment your topic idea below.
All findings will be shared here.
🚀 We’re launching our first non-profit research on startup challenges
We want the community to shape it
💡 If startups are your target customers, what do you most want to learn about their needs or pain points?
Comment your topic idea 👇
All findings will be shared here
🚀 We’re launching our first non-profit research on startup challenges
We want the community to shape it
💡 If startups are your target customers, what do you most want to learn about their needs or pain points?
Comment your topic idea 👇
All findings will be shared here