Thanks for joining us at a Get Started with Rovo session!
Here's a recap of the prompting framework we walked through together. Bookmark this, try the framework on your own, and join the discussion to share how you’re creating prompts that help you with your work!
The core idea we discussed during the live session is simple: what you put into the prompt usually decides whether the AI answer is useful or useless.
Think of this as ordering coffee. “Can I get a coffee?” triggers more questions, but “oat milk flat white, extra hot” gets you exactly what you want. Rovo works the same way: the more detail you give upfront, the less you fix or change afterward.
The framework
There’s a simple, 5-part framework for this called TCREI:
- Task: Tell Rovo exactly what to produce. (A step-by-step troubleshooting guide in bullet points.)
- Context: Add the details that make your situation specific. (Netgear Nighthawk router, two-story house, connection drops in the evenings.)
- Reference: Set constraints or point to sources. (Prioritize the most common causes first. If you need more details, ask.)
- Evaluate: Read the output critically. Does it actually solve the problem, or is it generic filler?
- Iterate: Send one targeted follow-up to reshape what's missing.
Here’s what a prompt created with TCREI can look like:
"You are a technical support agent. My home Wi-Fi is slow and keeps dropping connection, especially in the evenings. I use a Netgear Nighthawk router and live in a two-story house. Please provide a step-by-step troubleshooting guide in bullet points, prioritizing the most common causes first. If you need more details, let me know what to check."
Take a moment to question the answer 🧐
Most people stop here because the answer looks reasonable. But looking reasonable and being useful are not the same thing. Spend some time with the output. Do these troubleshooting steps actually make sense for this specific router? Is the tone right for the person reading it?
Now, move to the next part: Iterate.
This is where you can look at the gaps, fix what's off, or ask Rovo to rebuild this into a different format. Try something like:
"Turn this into a 5-question FAQ I can pin in a team channel so my teammates can reference it quickly."
You keep the exact same troubleshooting facts, but you’re just molding the first draft into the exact shape you need.
Try it on your next prompt
Next time you open Rovo Chat, take five seconds before hitting enter. Check for a clear Task, specific Context, and at least one Reference. When the answer comes back, don't accept the first draft. Evaluate it and then Iterate on the outcome.
🌟 Share a comment below with your own before-and-after experience. What was your initial prompt, and how did your follow-up iteration change the final result? That gap between the two is usually the most interesting part.
Keep learning
