Hacker News Reader

Comment by FunkyFreddy | Hacker News Reader

I am a product guy and was curious about a never ending probolem my girlfriend voiced to me.. the issue that lots of women struggle to make a decision on what to wear and not having confidence in knowing what looks good on them.

It starts with a colour analysis to give styling recommendation that I see becoming a styling assistant that can interact chat and voice based..

It can utilise AI and algorithm based product suggestions.

A super interesting problem to solve using genAI tools and new foundation models today. A clear value proposition and use case.

I need a full stack person to build this with me.

View on HN

What are you providing to the dev building this? Can't they just build it without you? You're on HackerNews...the home of self-agency...

Voice chat feels unnecessary and bordering on being predatory towards creating parasocial relationships with your 'advisor', imo.

How do you expect this to handle poor lighting vs different skin tones? Can you differentiate between a poorly lit room and a darker skin tone? To me, this would be the largest hurdle to clear

I am aware of the fact that I am on HackerNews ;) and thats why I am posting it here.

It is extremely interesting if you look at the current state of the art of AI tools to use and build on top of with this product.

The dev could become a co-founder or be compensated for the success of this.

Ultimately, I think the likelihood of succeeding is 10x with a good understanding of branding (ui/ux) - its not a tool for developers. It is a consumer product for a very different user.

And understanding the user is a challenge, if you are not the user yourself.

I realized my inital comment came off a little snarky as well, and added some feedback and some hurdles I, with my very little experience, forsee.

I want to add, I am NOT a full-stack dev...just killing some time before I leave the office.

Hope I didn't come off as too harsh!

Maybe a good example is this: https://elfnalysis.elfcosmetics.com/

So the tech for the picture to propose a colour type for in the analysis isnt so hard imo.

The elfnalysis example is very poorly executed thinking about the output that it generates.. there is really no value.

But then extending this into a better and more in depth analysis for more styling tips and feedback is where I would take this.

no worries :)

What would a dev provide to you? Can't you just build it yourself using AI coders? You're on HackerNews...the home of spamming advertisements for ChatGPT-wrapping SaaSs...

haha true.. well I think there are some complexities that I would like to collaborate on with someone with more technical knowledge and depth.

You can reach out to me in this regard.

How can I reach out to you for this?

Reply on news.ycombinator.com