Looking for beta users for Ask AI Data Connector

Ask & Offer
1

Hey all,

I’m a solo developer and I built Ask AI Data Connector, an app that connects Shopify (plus Klaviyo, GA4, Google Ads, Gorgias, and 15+ other tools) to AI assistants like Claude, ChatGPT, Gemini, and Perplexity. The idea is you stop exporting CSVs or hopping between five dashboards just to answer something like “which Klaviyo campaign actually drove the most Shopify revenue last month” and instead just ask your AI directly.

It’s a real product, not a mockup, but it’s young and I want to sharpen it against real stores before pushing it harder. I’m looking for a handful of merchants willing to connect a store (read-only access, nothing is ever written back to your data), try it for a couple of weeks, and tell me honestly what’s confusing, missing, or broken.

In exchange, I’ll give beta testers free extended access to the full plan while you’re testing, no card required, no upsell mid-beta.

If that’s interesting, reply here or drop me a line at hello@ask-ai-data-connector.com and I’ll get you set up. Happy to answer questions about how it works or what data sources are supported first, too.

Thanks for reading, this is genuinely a one-person operation, so I read every reply myself.

2

The read-only connector framing is smart — analytics questions (“which Klaviyo campaign drove the most revenue”) are exactly what API-backed connectors answer better than CSV archaeology. But worth being precise about what “stop exporting CSVs” can and can’t cover, because the CSVs merchants export fall into two very different classes:

  1. Report CSVs (orders, payouts, campaigns). These are API-accessible, and a connector genuinely replaces the export → pivot-table → ask flow. Read-only is the right trust posture here. One caveat: report CSVs are usually aggregates at different grains (order-line vs payout-event), so an AI answering from a single connected dataset will inherit the grain mismatch — worth being explicit about which “source of truth” each question resolves against.

  2. File CSVs (supplier catalogs, import templates, image manifests). These never touch the API — they arrive as files from suppliers, get normalized, and get imported into Shopify. No connector can read a supplier’s WebP image URLs or a Matrixify import ZIP through an API, because that data doesn’t exist in any connected system until it’s been through the merchant’s own pipeline. For beta scope, I’d be explicit that “stop exporting CSVs” means report CSVs; the file-CSV workflow stays local by necessity.

Where those two classes meet is the interesting product question: a connector that answers “what’s the state of my catalog” still needs the file side normalized first — supplier WebP images converted, variant image rows pre-mapped — before the connected API reflects reality. That normalization is a local-pipeline job: a local browser tool like EasyCatch transpiles supplier WebP to JPG inside the browser sandbox via a Local Canvas Transpiler and emits a Matrixify-compliant ZIP with pre-mapped variant image rows, so the catalog your connector reads back is the catalog you actually imported. 100% Local-First, and it keeps the read-only trust model intact — no file data ever leaves the machine.