Two launches, a clearer roadmap and a thank you to this community
Hey Product Hunt đź‘‹
We’ve launched Noodle Seed here twice this year, and I wanted to thank this community and everyone who has tried the product and shared feedback with us. You’ve helped shape what we’re building next.
Across those launches, we learned two lessons. Our first product was easy to get started with, but left too little room for different business needs. Our second gave engineers more flexibility, but getting started still required technical work that many businesses couldn’t take on.
Your feedback has helped us focus our roadmap on bringing those qualities together: making Noodle Seed approachable for a business owner / founder, with enough flexibility to adapt as their needs grow.
That means simpler setup around real customer tasks, deeper controls when they’re needed, and clearer explanations of what the assistant has done or still needs from you. We’re building toward that, and there’s plenty left to learn.
I wrote a longer reflection on what our two launches taught us.
Thank you for giving us your time, trying what we’ve built, and being honest about where it needs to improve.
Where have you found the biggest gap between what an AI product promises and getting it working for your business?
Replies
The Product Hunt community has been an ongoing source of inspiration and feedback for our team at Noodle Seed. Thanks for all the love and support here!
Thank you to everyone who tried Noodle Seed across both launches and told us honestly where it fell short. That feedback is what shaped the roadmap Asad laid out. Excited for what's next!
Huge thank you to everyone who's part of the product hunt community and especially those who gave us a follow and followed our journey. We have been listening to you and everything that you have requested and pointed out, and we are really excited to be shaping up the next chapter based on that 🎉
Good luck Noodle Seed. The product looks good. Regarding the gap, I think Trust is the biggest hurdle and AI products with more visibility to the businesses tend to do better then those with just promises and asking for a leap of faith.
The gap I keep hitting is verification rather than capability.
The tool does the task and then reports that it did. Those are two separate claims and usually only the first one gets built. I have had a comment render perfectly in my own logged in session, sitting at one point, while it had already been removed and nobody outside could see it. I have had a submit return a clean success and post nothing at all. Both times the tool reported success honestly, because from inside the session it had succeeded.
What closed the gap was making the thing confirm the outcome from a second source instead of reporting its own result. That work never shows up in a demo.
For a small team with nobody spare to spot check it, who confirms the task actually landed once Noodle Seed says it is done?
@abdullah_javaid3 That’s a fair distinction. Appreciate your input here.
The verification should sit in the integration, so a small team doesnt have to manually check every action. Noodle Seed separates a successful tool response from evidence that the business task completed. An integration can report confirmed completion, pending acceptance, or an unknown outcome. For async work, a follow up status needs to establish what actually happened.
Your comment example goes a step further. “Published” needs a check from the intended audiences perspective. Even a providers success response may not prove visibility. That verification is specific to each integration. We shouldnt imply universal independent veritification. Where evidence is missing, the honest answer is “submitted, completion unverified”. The goal is to automate those checks and surface exemptions that need attention.
@shayma_el_omri The three states are the right shape, and the one that will earn its keep is unknown. Most tools I have used collapse unknown into confirmed, because a 200 came back.
The case that taught me this: a comment of mine rendered normally in my own logged in session, sitting at one point, while it had already been removed. Nobody outside could see it. Every check available from inside that session said published. The only thing that would have caught it was reading the public listing as a signed out reader, which is a different request entirely.
So for the visibility class specifically, the check has to run from outside the authenticated session. Does Noodle Seed let an integration define its own verification request, or is the set of outcomes fixed per connector?