Skip to main content
Rilo gets more useful the longer you work with it. After each task, Rilo retains what it learned about how you like things done — and applies that to future tasks automatically.

​
What Rilo Remembers

Your preferences How you like results formatted, what level of detail you want, which tools you prefer for what kinds of tasks. Your processes If Rilo figures out the right sequence of steps to accomplish a recurring task, it remembers that approach and applies it next time. Context about your organization Who your key accounts are, what tools your team uses, terminology specific to your industry or company. What worked If a particular approach produced good results, Rilo notes it. If a task needed correction, Rilo factors that in going forward.

​
How Memory is Shared

Memory is shared across your organization — not per person. If your colleague works with Rilo and establishes that the team prefers reports in a particular format, Rilo will apply that for you too. This means Rilo gets better for your whole team over time, not just the individuals who use it most.

​
Setting Preferences Explicitly

You don’t need to wait for Rilo to infer a preference. You can state them directly in any message: 
@Rilo My preference going forward: when you write outreach emails, keep them
under 80 words and never mention pricing in a first touch.

@Rilo For any research task, I always want a LinkedIn URL included for the
key contacts, not just their names.

@Rilo Going forward, format all summary deliverables as bullet points, not
paragraphs. I read them faster that way.
Rilo will confirm it’s noted and apply the preference to future work.

​
Correcting Rilo

If Rilo does something the wrong way, tell it — and tell it what you’d have preferred: 
@Rilo This is too long. In the future, keep research summaries to half a page
maximum unless I ask for more.

@Rilo I didn't want you to email them directly — I just wanted a draft for me to
review. Going forward, always check before sending anything externally.
Corrections work the same way as preferences: Rilo notes them and adjusts.

​
Memory Within a Task

Rilo also carries context within a single task session. If you add context mid-task or redirect Rilo partway through, it incorporates that and continues — it doesn’t restart from scratch. See Following Up Mid-Task for how to add context, change direction, or stop a task while it’s running.