I built a business-memory AI that watches your shop's database and tells you what's wrong A retail shop owner built Cuedis, a business-memory AI that monitors a shop's POS, ERP or accounting database (SQL Server or MySQL) and sends a plain-language morning briefing via Telegram flagging anomalies such as a quiet supplier, a stalled till or a sudden sales spike. The tool runs read-only on the shop's own PC so data never leaves the store, learns each business's normal patterns, and answers natural-language questions in about a second; the developer says it matched their own store's POS reports to the cent. A free version for one database is available at cuedis.com, and the product launched on Product Hunt. I run a retail shop. Every morning I used to open the POS reports and dig through yesterday's numbers to find out what went wrong. Most mornings, nothing had. Some mornings, something had — a supplier going quiet, a till that stopped ringing, one item suddenly selling 10x — and I'd only find it days later, after it had already cost me money. So I built Cuedis: a business-memory AI that watches your shop's database and tells you what's wrong, in plain words, before you go looking for it. Cuedis sits beside your POS, ERP or accounting database SQL Server / MySQL . It's read-only — it never writes to your data, and it runs on the shop's own PC, so the data never leaves the store. It learns what's normal for your business, then every morning it sends a short briefing on Telegram: And you can just ask it questions in plain words — "which supplier has been late most this month?" — and get an answer in about a second. On my own store it matched the POS reports to the cent. The thing I cared about most was that it had to work for a shop owner, not a data analyst: no dashboards to learn, no queries to write. A short message on your phone each morning, and an alert the same day something changes. It's free for one database — register and download the free version at https://cuedis.com https://cuedis.com It's also live on Product Hunt today: https://www.producthunt.com/products/cuedis https://www.producthunt.com/products/cuedis Honest feedback welcome — especially from anyone running a shop on a local SQL database. What would you want it to watch for first?