> ## Documentation Index > Fetch the complete documentation index at: https://docs.slingso.com/llms.txt > Use this file to discover all available pages before exploring further. # Integrations > Slingso works from what human and AI buyers actually see and do — a deliberate design choice, and the whole idea. Slingso works from what your buyers see and do — the people shopping, and the AI agents shopping for them. That's by design, and for a good reason: mirror the buyer closely enough and you're looking at the exact thing that decides the sale. Deeper integrations are on the roadmap and will smooth a few edges, but they were never what makes Slingso work — this principle is. ## Slingso stands where your buyer stands The AI Shopping Channel is decided in the open — on the public web, in the answer an AI assistant reads back to a shopper. So that's where Slingso works: from the same live pages, prices, reviews, and answers a buyer — or the AI shopping on their behalf — actually encounters. It's the only view that decides a sale. Your Shopify dashboard can say **"in stock."** But if the live page a buyer's AI reads comes across as unavailable, the AI moves on and buys elsewhere. It decides from the outside, on what it can see — not on what your admin knows. An integration would show you your *internal* truth — your specs, your stock, your prices. You already have that. The truth that wins or loses the sale is the one the outside world *perceives* instead, and that only shows up from the outside. ## That gap is the entire job Everything Slingso measures is the distance between the story you intend and the one AI actually tells shoppers — your [Narrative Accuracy](/guide/your-two-scores). Your internal systems hold the intended version; you don't need a tool to read your own database back to you. The part that costs you money is what gets perceived instead — a missing spec, a stale price on a marketplace, a bestseller that reads as sold out. Slingso watches your public pages for exactly these, and flags a product going dark while AI is still recommending it. ## Tune what the whole internet sees The lever that moves AI's recommendation is your public presence — the pages, prices, reviews, and lists that both human shoppers and their AI agents read. Tune that, and the recommendation follows. It's the surface Slingso works on and helps you improve: the one where the buying decision is actually made. Your own systems still run your business day to day — they're just not where AI forms its view of you, so that's where Slingso spends its attention on your behalf. ## A buyer's AI shops from the outside, too As shopping moves to AI agents that browse and buy for people, this only gets truer. A buyer's agent doesn't log into your store or read your product feed — it visits the same pages, marketplaces, and reviews any shopper would. The surface Slingso works on is exactly the surface those agents use. There's nothing to integrate, because the buyer never integrates with you either. ## And it runs on about fifteen minutes a week Because the agent does the watching, diagnosing, and drafting, your part stays small: approve what makes sense, make the real-world calls only you can, and publish the content by hand — a copy, a paste, a link back. No data pipes to wire up, no dashboards to maintain. Roughly fifteen minutes a week, and the channel keeps moving. Integrations like one-click publishing are coming, and they'll save you those few minutes. They'll never be *why* Slingso works — it works because it sees what your buyer sees. Narrative Accuracy is the gap this page is about — measured.