Plans 💳 : Everyone
User Permissions 👥: Everyone
Curious to know more about how Memory AI works in the background? Below you'll find answers to some of the questions we get asked the most by users interested in how the AI works.
How do I train Memory AI?
Once you've installed Memory and connected any apps you want with Timely, you can start training the AI by logging time with Memory using either the timeline or the list view. The AI will start learning from your very first logged entry.
Why do I need to log from Memory to get the AI working?
Even though all of the data is available in Timely, Memory AI needs input from you to get started. That input comes in the form of the hours you log from Memory, which tells the AI what projects you want those memories to be associated with. For those to leverage Timely's Tags feature for additional specificity, Memory AI will learn from how you tag your time entries over time and suggest a specific Tag in your AI drafts.
The results of that input are suggestions made by Memory AI that you can adjust then approve or reject. What really helps train the AI is not only the input you provide from Memory but also adjusting and approving the suggestions the AI makes so it can improve and make more accurate suggestions going forward.
How long does it take Memory AI to work?
There is no set amount of time before Memory AI starts working. Memory AI will start learning your logging behavior as soon as you’ve logged your memories for the first time. The more you log, the faster you will receive suggestions. If you know that Excel file always belongs to your "Accounting" project—log it and log it often! When it comes to logging memories, our AI developers say: "quality over quantity, but really, both!"
Does Memory AI work the same for everyone?
Memory AI is unique to every account and every user. For instance, if you have a personal Timely workspace and one you use with your team, each workspace will have its own unique Memory AI.
What should I do if a memory is logged to the wrong project?
First, un-log the memory from the project it's not supposed to be logged a part of, then log it to the right project—easy peasy! Very straightforward but don't ignore that memory as that can throw off the AI when it tries to match that memory to the right project going forward.
What happens when the suggested Tag doesn't match the memories?
Go ahead and update the AI draft to remove the incorrect Tag and assign what you feel works best. Keep in mind that, for now, Memory AI can only learn from those time entries you create that contain a single specific Tag. It will then do it's best to suggest a relevant Tag in the resulting AI drafts.
As you continue to correct your AI drafts and log your time normally, Memory AI will continue to learn from you and those suggestions will improve over time.
Do I need to Tag my time entries in order for Memory AI to learn from them?
No - Memory AI will continue to take your un-tagged time entries into consideration, along side any time entries that contain a single specific tag. Tagging is always optional!
What happens if I’m logging memories properly and things still don’t change?
If you notice that Memory AI is still offering suggestions that aren't right, there are a few things you can do. One option is to go back using your timeline and make sure memories are logged to the right projects, fixing any memories that might have been mis-logged along the way.
Another option is to continue logging memories to their projects going forward and Memory AI will catch-up eventually. Since every user's Memory AI is unique, we can't say for sure how long this method will take so the option you pick will ultimately depend on how many memories need to be fixed in the past.
Does Memory recognize my file names when making suggestions?
If the title of the browser window/tab or app window matches a project you've logged similar memories to previously, Memory AI will notice that while making suggestions. Keep in mind that Memory AI won't make that connection unless it has been trained first. Though not an exhaustive list, this will also happen with things like email addresses, domain names, URLs and file paths, for example.
What won’t Memory recognize?
Raw data from Memory doesn't contain anything other than details from the active browser window/tab or app window. As a result, Memory AI isn't able to draw connections between your logging patterns and anything that isn't uploaded to Memory, such as content from emails, files, websites or your computer.
Feel free to check out this article if you want to know more about what data Memory captures.
Can Memory read notes in Timely entries?
Notes in your logged or planned time entries in Timely do not influence Memory AI.
What happens if I delete memories?
Deleting memories doesn't influence Memory AI. If you delete a memory from your timeline or list view, it just goes away—that's it!
Should I ignore memories? Is it good for Memory AI?
Ignoring memories can be good for Memory AI but it is something we advise you to be careful with. It's another form of input for Memory AI and it can also help you keep a clean timeline but if there's any opportunity at all to log a memory to a project, we recommend that instead of ignoring the memory altogether. Once a memory is ignored, Memory will continue to ignore this memory unless you restore it.
What should I do for memories that are non-work related?
Sometimes there's no avoiding that funny cat video or digging for the perfect meme that sums up your mood. With these memories that you won't need to log to a project, you can either ignore them or delete them in order to clean up your timeline.
Can I disable Memory AI?
You can disable Memory AI by going to the Hours tab and clicking the brain icon near the top of the screen. From the prompt, use the "Timely AI tools" toggle to switch this feature on or off.
Why isn’t Memory AI logging my time for me?
Instead of logging your time for you, Memory AI drafts time entries using your activity and logging patterns. They need your approval to appear on your public time sheet.