I’m 17 and I built an Agentic AI to kill the 15-minute Shopify listing grind

The 15-minute Shopify listing grind is real and you picked a wedge that is

already paying. Few questions from someone shipping in a different vertical:

1. How are you handling the source-of-truth problem when the merchant has

listings in 5 places (Shopify, Etsy, Amazon, Instagram shop, TikTok shop)

that all need to stay synced after edits?

2. What is your defensibility once Shopify itself ships their own AI listing

generator? They already have one in beta. The opening might close within

12 months.

3. At your age, the unfair advantage you have is time and zero context

switching cost. Use it. Most adult founders are envious of exactly that.

Keep shipping. The product looks tight

Solving a specific painful workflow is exactly the right

approach. Execution speed matters more than age.

What's been your biggest distribution challenge so far?

17 and building agentic AI? Impressive. 'Killing the grind' is exactly where the value is right now.

One thing I've noticed with these specialized agents is that the 'hidden grind' often moves from the user to the developer. Managing the underlying AI accounts, rotation, and limits to keep the agent running 24/7 is a job in itself. That's why we built AiKey — to automate that backend chaos so we can focus on the actual agent logic.

Are you running this on a single API key, or are you already looking into account pooling to handle scale?

pretty impresive, looks really clean, best of luck with your project

17 and already solving a real operational pain — respect. The 15-second pitch is sharp and the demo does the job.

One thing worth pressure-testing: the $1.99 price point might actually hurt you. Merchants who list at volume have budgets — too cheap signals "experiment" rather than "tool I depend on." Worth testing a higher anchor early while you have the "free access" hook to lower friction.

Biggest moat question: what stops Shopify from building this natively? Your answer to that shapes everything. 👀

The Shopify App Store has its own gravity, that is the channel that will make or break this. Two things matter way more than people outside it realize.

Install-to-paid conversion in the first 14 days will be determined by your onboarding video and your first 5 reviews, not your pricing. A 90-second demo embedded in the app listing is worth more than any third price tier you could split test.

Also, the people who feel listing pain most acutely are not high-frequency listers, they are dropshippers and resellers importing 50 to 500 SKUs at a time from AliExpress, CJ, or Spocket. The 10-minute pain in their world is a 5-hour pain. Your starter tier of 100 generations does not cover one batch import for that ICP. Consider a one-time bulk credit pack ($9.99 for 500 generations) for new store launches. That is an impulse purchase, not a subscription decision.

Building this at 17 is impressive, but the angle gets you attention not retention. Strip it from the marketing once you have 50 paying customers, the product has to stand on its own.

Wow

15 mins per listing adds up fast - 100 products = 25 hours of soul-crushing work. I've seen merchants quit after 50 listings because the grind didn't scale. curious: do you flag low-confidence outputs when supplier data is genuinely terrible?

Really impressive you're building this at 17. The problem you're solving is real—listing creation is a huge time sink for Shopify sellers. My advice: don't try to boil the ocean. Pick one type of "messy data" (scanned spec sheets, CSV exports from suppliers, or handwritten notes) and make Filleo absolutely perfect at handling just that. Nail one use case first, then expand. Good luck.

This is seriously impressive. Focusing on messy real-world supplier data instead of just “AI for the sake of AI.” is a very good product direction! Keep up the good work!

Have u tried yearly plan? can u share some stats on subscription choices?

Cool

Honestly, the messy supplier-data problem is very real — especially for stores importing lots of products from different vendors. Cleaning titles, descriptions, tags, and formatting manually is one of those repetitive tasks that eats a surprising amount of time.

The $1.99 entry point feels reasonable for getting people to test it, especially if the output quality is strong and saves real editing time. I think the biggest factor will be how well it handles inconsistent or low-quality source data in real-world cases.

Also impressive project to be building at 17 🚀

Honestly, the messy supplier-data problem is very real especially for stores importing lots of products from different vendors. Cleaning titles, descriptions, tags, and formatting manually is one of those repetitive tasks that eats a surprising amount of time.

The $1.99 entry point feels reasonable for getting people to test it, especially if the output quality is strong and saves real editing time. I think the biggest factor will be how well it handles inconsistent or low-quality source data in real-world cases.

Also impressive project to be building at 17

Great work buddy, keep the hustle

This is actually a really cool idea — the Shopify listing pain is way more real than people think, especially when you’re scaling catalogs manually.

What I like here is how practical it is. Taking messy supplier data and turning it into a clean, publish-ready listing is exactly the kind of repetitive work merchants don’t want to think about.

At this stage, I wouldn’t worry about pricing yet. I’d just try to answer one simple question:
👉 does this reliably save time without messing up the listing quality?

If yes, everything else gets easier later.

Also, building something like this at 17 is impressive — but the real test will be getting a few people to actually use it more than once, not just try it out.

Curious how it handles messy or incomplete supplier data in practice 👀

Impressive for 17. How do you handle edge cases in the ad generation? Real‑world testing must be tough.

This is actually a very real pain point.

Most people underestimate how much time product listing takes until they start doing it daily. 10–20 minutes per product sounds small, but at scale it's huge.

$1.99 feels like an easy yes if the output quality is consistent.

good i like it

Smart move targeting high-frequency listers first. That's the group where the time savings compound fastest, and they're the ones who'll actually feel the pain enough to pay.

One thing worth thinking about early: the quality gap between what your AI generates and what a merchant would write manually. For commodity products the bar is low so automation wins easily. But for branded or premium products, merchants get really particular about tone and description style. Might be worth letting users set a "brand voice" template so the output doesn't feel generic.

At $1.99 you're basically removing price as a decision factor which is great for early traction. Just make sure you're tracking which features drive retention vs which drive signups, they're usually different things.

Really impressive for a first build. The fact that you went straight to a real workflow problem instead of building something cool but useless puts you ahead of most builders I've seen, regardless of age.