Most apps on your phone could be an agent you chat with or talk to. But where are they?I put together a phonebook of these AI bots. Still scraping, will add more as we go.Know one that should be listed? There's a suggest form.Stack is Astro + Cloudflare to keep it simple. Cala.ai for scraping (free credits from a hackathon)
I think that's because time tracking is hard and annoying, that's why I made it "set and forget" instead of requiring a new habit.
https://donethat.ai/blog/dogfooding-donethat
Hey HN!I wanted a time tracker accurate enough to be useful and easy enough to not require a new habit while protecting your privacy.Most trackers (RescueTime etc) only use activity logs and that isn't enough to be very useful. DoneThat uses LLMs to parse your screenshots, which makes it much more accurate but also requires much higher privacy standards, so let's start with those:-...
Right now working on framing this as an RL problem to better predict when nudging is actually successful and what kind of wording works best given the user data. Then applying the same logic to onboarding emails etc.
A tool to estimate if you should vibe an automation/app or just buy/delegate/grind instead
https://www.euzoia.org/p/finally-stick-to-your-habits
Side project that'll hopefully raise some money for charities!Behavior change research shows that putting money on the line makes it much more likely that you'll stick to your habits. It's not being used much though because most people don't want to put money on the line hahaWe think part of the reason is that your money is traditionally going to the startups (eg Beeminder, Forefeit). We solve...
So spyware in the sense of getting information without the employee knowing would be impossible and not something I’d ever want to do.
It does enable transparency on a very abstracted level: your team could see a six bullet point summary of your day if you opt in. I believe this kind of transparency can actually help more teams go remote, cut down on sync meetings, etc.
I’m currently experimenting with a feature that shows relative time spent only, not absolute - so e.g. 30% on project X, 20% on admin, etc. That could be the sweet spot on visibility vs privacy.
Full project: https://euzoia.org
Tried to be super low-tech: Notion, super.so, Spotify creators, riverside.
Now thinking of building an email-based agent for behaviour change accountability. Would love any pointers to good UX for email-based AI assistants.
An AI based time tracker: reconstructs your day from whatever it sees you doing. Screenshot based but never stores them.
The same tech stack is pretty easily adaptable to openclaw tracking. If anybody would like to try, DM
Also looking into AI based security tools for monitoring security of DoneThat. Thinking of using zeropath would love to hear if people tried them / have other suggestions
https://euzoia.substack.com The concept: https://euzoia.org
Tried to do this as low tech as possible, so website is just an off the shelf notion wrapper
Currently experimenting with a proactive agent, Don, that pops up like clippy and also works over email.
https://donethat.ai All the data security measures here: https://donethat.ai/data And other tools out there: https://donethat.ai/compare
With lot's of built-in data privacy safeguards https://donethat.ai/data
Also made an overview of similar tools out there https://donethat.ai/compare
Recently broke on Linux with a Wayland security update, working on a fix! Using Electron for cross-platform.
https://donethat.ai Passively processing screenshots is obviously pretty sensitive, it has an option to bring your own (local or remote) LLM, otherwise I process with gemini and never store any data.
It's in beta right now so if you want to try it you have to enable "proactive chat" in settings.
I also made a list of similar tools out there: https://donethat.ai/compare
Obviously this is quite sensitive data so architected it to never store raw data, allow for bring-your-own-key, and even in team settings be fully private by default, everybody keeps control of all their results.
Started about six months ago, have some first users, and always looking for feedback!
> A local screen recording app but it uses local models to create detailed semantic summaries of what I’m doing each day on my computer. This should then be provided as context for a chat app. I want to ask things like “Who did I forget to respond to yesterday?” I've been using Rewind for a year now, and it's nowhere near as useful as it should be.
I am building something like this but unfortunately not local because for most people's machines local LLMs are just not powerful enough or would take too much drain on battery. Work in progress, always curious for feedback! https://donethat.ai
If you want fully local, somebody did a post on HN on something related recently: https://news.ycombinator.com/item?id=45361268
The concept is pretty simple: It takes screenshots every few minutes and reconstructs a calendar of what actually happened. On top of that features to set goals for how much time to spend on topics and a social feed.
This is obviously super sensitive data. I built it in a way that it never stores raw data and you can also bring your own Gemini key, plus a lot more privacy features. Tried to make it fully local but found that it takes too much energy and only works on high end machines.
I have a few first users, iterating with them, and right now exploring if it’s possible to make this work in team settings without sacrificing privacy.