Conception
AlertChecker was created to help fill a perceived gap with AI chatbots: They only respond when prompted and won't spontaneously tell you anything.
AlertChecker began with a familiar frustration: I knew what I wanted to know but didn't know when it would change. I could look up a sold-out product, search for a room price, check for an event announcement, or ask about a weather warning. But the next step was usually to move on to something else and remember to try again later.
Answers vs. follow-up
Search engines, websites, and chatbots are generally good at providing current information. The normal flow is user-invoked: you open a tool, type a question, and usually receive a reply that is correct at that time. That's great if all you need is instant research, but it won't come back and tell you when the situation has changed.
If you ask a chatbot whether concert tickets are on sale, it can investigate the current position. But it won't proactively tell you next week when new tickets have been issued. The same is true when you ask about an item being in stock or a holiday price falling. The onus is on you to keep checking.
The missing piece
AlertChecker changes the shape of what needs to be asked. Instead of repeatedly saying “Is it true yet?”, you write what you'd like to know about when it happens: “x is in stock”, “The price of y is below £120”, or “Registration for z is open”. It then monitors that condition over time and sends an email when the answer has changed.
This sits between one-off research and conventional page monitors. You don't necessarily need to know which webpage will carry the final answer, nor to worry about locating elements on pages, dealing with changes in content etc. - AI handles all that. The main thing you need to provide is the condition that must be met to trigger an alert.
An AI assistant companion
AI assistants make it easier to investigate and phrase a useful alert. AlertChecker adds the part those conversations are not designed for: waiting and following up. If you set up an MCP connection, the assistant can even create an alert whenever you request it in a normal chatbot conversation.
The aim
The simple aim is to replace repeated checking by a human with one clear statement and a timely notification. For questions that would otherwise stay in the back of your mind, an alert can make the web work on your schedule instead of the other way round.