Fabrice Salah

Super useful tool! As someone who appreciates optimization problems, I love how Visual Cut List makes material cutting so intuitive — drag, rotate, and instantly see your material usage and costs. The 1px = 1mm approach is clever and keeps things simple. Great for anyone working with wood, metal or textiles who wants to minimize waste. Nice launch! 🪵

Love the insight behind this — chat really is the wrong interface for managing agents. As a CPO, I've seen my team struggle with keeping track of what AI agents are actually doing across multiple tasks. Giving them a proper task board like you'd use with any teammate just makes sense. The Notion integration angle is clever too. Congrats on the launch, excited to try it out with our team! 🚀

The free tier being limited to 20 seconds makes it pretty hard to properly evaluate the tool, especially when the maker also mentioned AI still struggles to animate product screens. If you’re a SaaS wanting to build explainers, it feels like the core use case isn’t fully there yet.

Only OpenAI and Anthropic protocols supported so if you need Gemini or Mistral youre already adding another gateway, hard to really call that a unified API with two providers

The maker confirmed it’s online-only, but the promise is “anytime you need it.” What happens if someone is in a panic situation on a flight or somewhere with no signal — is there any fallback or is Anna simply unavailable?

Gmail only for email automation + no ICP builder wizard yet feels like a rough combo for the "founder-friendly" positioning. Most early founders dont even know their ICP clearly, thats kind of who you need to help the most.

The platform and the library of automation ideas are both listed as "coming soon" so right now the core deliverable is a Discord server and weekly calls. For someone joining today at $299, how structured is the onboarding to actually find the right configs and channels instead of scrolling through months of chat history?

The 3-step flow sounds clean but what happens when the agent generates a form the user submits? Where does that submitted data actually go, and can you wire up downstream actions from it or is it just stored somewhere?

I'm trying to understand the workflow here. So a broadcaster plugs in a live stream via the web platform or API, picks target languages, and the dubbed audio goes out in real time? What does the experience look like on the viewer side, do they choose a language before the stream starts or can they switch mid-stream?

So the auto plug triggers based on likes thresholds and time right? How granular can you get with that, like can you set different plugs for different types of posts?

I get the alt-tab pain point, but how does visual editing actually work inside the dock? Do you edit live and it pushes to Cursor, or is it more of a preview before sending?

Quick question Michael, when a designer redesigns a nav bar visually and pushes a PR through Inspector, is there a review step before merge or does it go straight to the repo?