Careersy AI: added Deep Research. Also, it now forgets things on a timer.
That's the whole entry, if we wrote it the way software usually writes these. One line, a feature shipped, and you'd skim past it in four seconds without ever learning that the strangest part is the forgetting, and that the forgetting is the point.
We're not doing that. This is The Debrief: what changed at Careersy AI, and the real moment that made us change it. Not a changelog. An entry only earns its place here if it traces back to something that went wrong: someone I watched get hurt by the old way, or the product caught doing something it shouldn't. No traceable failure, no entry, however proud the product team is of it.
This edition turned out to be about memory. Not on purpose. We looked up at the end of the week and realised half of what we'd built came down to one idea: software should remember what you'd hate to redo, and forget what you'd hate to be reminded of.
Careersy AI: Deep Research runs live web research and keeps it in a Research Library.
I once watched a candidate quote a company fact that was eighteen months out of date, to a hiring manager who had lived through exactly why it changed. The room went quiet in a specific way. She had prepped. She'd just prepped once, weeks earlier, and the company moved on without telling her.
Deep Research is the fix for the first half of that. Press the binoculars in the chat bar and ask about a company, a role, or the person interviewing you. Careersy AI goes out to the live web and comes back with a sourced, chaptered report: where they're heading, how they actually hire, what their people have said in the last six months, where they stand on AI.
The Research Library is the fix for the second half. It keeps every report, privately, so when you open a fresh chat three days later to prep, your coach already knows what you found and the date you found it. No re-pasting, no lost tab. Your research compounds instead of evaporating the moment you close the window.
And then it forgets it, on purpose, on a timer, and here is the reason. Stale prep is worse than no prep, because no prep at least knows it's guessing. So research about a person expires after 30 days, everything else after 90. Careersy will cheerfully forget things on your behalf, which is a strange sentence to be proud of, and here we are.
Read the full Deep Research story →Careersy AI: added a Lite plan. A$25 a month, 25 credits, every tool.
Here's the failure this one traces to: the price was a wall for exactly the people who needed the help most. Not everyone can put fifty dollars a month toward a subscription right now, and that has nothing to do with how much they need it. You might be a graduate with no income yet. You might be months into a search with the savings gone and nothing spare for anything that is not rent. Both are ordinary places to be in this market, and neither should decide whether you get good help.
So we built a smaller door. Careersy AI Lite is 25 credits a month for A$25. Every tool, nothing locked, the same coaching and CV work and live job matching as any other plan. It is real value at a fair price, an actual entry point, not a stripped-back trial. Start here, and when things pick up, upgrading is one click and only charges the difference.
We wanted a way in that fits any budget. If money is the reason you have been putting this off, or the reason you drifted away when the free credits ran out, this is the door built for you.
See the plans →Careersy AI: memory is now editable in plain language, under Privacy and memory.
Careersy remembers what you tell it, your work, your goals, the way you like to operate, so its advice fits you and not a composite stranger. Useful, right up until you wanted to check what it had remembered, at which point you were reading a list that looked like a database export and changing things one row at a time.
Now you just say it. Open Profile, then Privacy and memory. There's a plain-English read of what Careersy understands about you at the top, and under it a box: "Remember I'm targeting Staff Engineer roles." "Forget the bit about Google." Ask it to forget something and it shows you exactly which facts are about to go, before they go, which is more notice than most people extend before deleting each other's numbers.
Careersy AI: dictation now removes the ums and false starts, and nothing else.
Speak into Careersy and your words now land as clean, polished text. The ums and the false starts come out; your facts stay exactly as you said them. It fixes how you said it, not what you said. We were very firm with it about that boundary.
Not everything traces to a coaching room. Some of it traces to us watching the product do something wrong and fixing it before more people hit it.
The "New" tags kept coming back. The "New" badges on things like CV Canvas and the What's new list lit up again on every single login, even after you'd already looked. A New badge that resets its own definition of new. We caught this one ourselves, largely because it was annoying us too. Fixed now: once you've opened something, its tag stays gone, so "New" means new to you. Nothing to redo.
One is on the way. We've lined up a provider to help us make short video tutorials, so you can watch how a feature works instead of reverse-engineering it from a paragraph. More when they land.
A normal changelog is written for the company: proof of velocity, aimed at people who already trust you and just want confirmation you're still shipping. That's not you.
Here's what actually matters, even though it won't get you a callback this week. You can't watch software get built, so you have no way to know whether the people behind it are watching what happens to someone like you, or watching a growth chart. Most roadmaps get built from a competitor's last release. Ours gets built inside a Careersy coaching session: a candidate quoting a stale fact into a silent room, a graduate who wanted the help and could not reach the price. The coaching is where we watch what actually breaks for people. The product is what we build back from it.
If there is a feature you keep wishing Careersy had, tell us in the app. We read every one, and the roadmap gets shaped by exactly that.
Don't take my word for it. Pick the company you're closest to interviewing with and run Deep Research on it this week. It comes back sourced, so you can check it rather than trust it, and if it surfaces nothing you didn't already know, that's data too. It's free to start, three credits, no card.
It runs live web research on a company, a role, or an interviewer and returns a sourced, chaptered report: the company's direction, how it hires, recent employee sentiment, and where it stands on AI. Press the binoculars in the chat bar to run it.
Yes. Every report is saved to your Research Library, private to you, never pooled or used for anyone else. It follows your memory settings, and research expires on a timer so your prep is never stale: 30 days for research about a person, 90 for everything else.
A lower-cost plan: 25 credits a month for A$25, with every tool included and nothing locked. It's a genuine entry point for graduates and anyone on a tight budget, not a stripped-back trial. You can upgrade any time and only pay the difference.
Yes. Under Profile, then Privacy and memory, you can read what it understands about you and tell it in plain language what to remember or forget. It shows you exactly which facts will go before it removes them.
Yes. Three credits to start, no card required. Every tool is included from the first paid tier up, and Lite keeps the full toolset at a lower monthly credit count.
Everyone gives you job-search tactics. Nobody warns you what three months of silence does to your head. Three concrete moves, from someone who read the applications, to take back the part the market works hardest to steal: a bit of control.
5 min read
Careersy AI can now run deep web research on a company, a role, or an interviewer, and it remembers what it found. Days later, your interview prep already knows.
6 min read
