# Obi Demo - 2min Cut ⚡

> **About Cor & Obi:** This is a short preview of Obi, Cor’s voice AI onboarding agent. It’s built to train and activate users at scale while reducing time spent on repetitive training sessions. 

**Call to Action:**[Book a demo or try Obi yourself at www.getcor.ai](https://www.getcor.ai)

---

## ⏱️ Video Chapters
* **0:00** - Intro
* **0:05** - Screen Share
* **0:15** - Guidance
* **0:45** - Recommendations
* **1:45** - Guardrails
* **2:00** - Validation
* **2:20** - Outro

---

## 📝 Demo Transcript

*Context: In this video demo, a user is walking through the setup of their own AI agent on the Cor platform. They are being actively guided in real-time by "Obi", Cor's AI voice onboarding assistant, which is able to see the user's screen and provide contextual instructions.*

**[0:00] Human User:** 
Hey, just a quick demo to show you how we use Obi to run our own onboardings.

**[0:04] Obi (AI Agent):** 
To better help you, I'll need you to share your screen. Please click on the screen share button in the control bar at the bottom of the Obi assistant window, and then share the window where you have Cor open.

**[0:15] Human User:** 
Yep, you should be able to see it now.

**[0:18] Obi (AI Agent):** 
Awesome, I can see your screen now. Next, go ahead and click on the "Agent" tab in the left-hand navigation menu, then click the "Edit Agent" button in the top right corner of the screen. This will let you configure your agent.

**[0:30] Obi (AI Agent):** 
In the configurations tab, you can change the agent's name, title, and description. These are all visible to your end-users. The title should describe what the agent is for, and the description gives a bit more context. Go ahead and make any changes you want here. Just let me know when you're done.

**[0:47] Obi (AI Agent):** 
Great. Now, could you tell me a bit more about your users? For example, do they often work in places with a lot of background noise like busy offices or outdoors? Are they in privacy-sensitive industries like finance or healthcare? And do your internal policies require session recordings to be deleted after a certain time?

**[1:06] Human User:** 
So our typical users are the employees of software companies and they do work in offices, but they're not that busy, typically they have a quiet space, so that's not an issue.

**[1:18] Obi (AI Agent):** 
Perfect, thanks for clarifying. Based on what you've told me, I'd suggest leaving the "thinking sound" and "interruptions" settings at their defaults, since your users are in quiet environments and interruptions can make the conversation feel more natural. For compliance, you'll want to turn on "Automatically Delete Recordings" and set it to delete after 30 days. You don't need to enable privacy mode unless you start handling sensitive personal info in the future. Go ahead and make those adjustments and let me know when you're ready for the next step.

**[1:47] Obi (AI Agent):** 
If you want to add more guardrails, you can click the plus "Add Rule" button. In the new rule form, just enter a title and a description to explain the guideline for your agent. You can add as many as you need. Let me know if you want to try adding one, or if you're good to move on.

**[2:01] Human User:** 
I've added one for pricing. Can you have a look and just tell me if it's correct?

**[2:06] Obi (AI Agent):** 
Let me take a look at your screen to check your pricing guardrail... I can see your pricing guardrail, and it looks good. The title "Pricing" and the description "Don't discuss pricing" are both clear and direct. That should do the trick—your agent will know not to touch the topic of pricing with a 10-foot pole.

**[2:19] Obi (AI Agent):** 
Thank you for your time. That wraps up the session. Honestly, you made configuring an agent feel like a walk in the park.

---

## 🔗 Resources
* [User Research](#) *(Replace with actual URL)*
* [Pricing](#) *(Replace with actual URL)*