Challenge:
The idea originated from the need to streamline the pre-sales process and be able to respond faster, from the moment an opportunity is identified to when the proposal is presented to the client, while always meeting the company’s internal standards.
Before getting started:
The first step is to ensure the agent will have strong information sources. That’s why we created two Confluence spaces: one to gather the requirements from sales meetings with clients, and another to centralize the details and characteristics that all pre-sales must meet within the company. For example, official structure, legal terms, rates, and examples of previous proposals.
Agent creation:
From the Rovo Studio module, I create a new agent. At that point, a text box opens up, very similar to the Rovo chat interface.
This is where you insert the prompt to define your agent. In this technote, Atlassian explains the best practices for crafting a good prompt. For this specific pre-sales agent, I’m sharing a snippet of the prompt I used:
Once you’ve told your Rovo agent what it needs to do, the next steps are the following:
At this point, our Pre-Sales Assistant is ready to use. However, we’re going to take it one step further: directly from Rovo Studio, we’ll create an automation so that, when a pre-sales request is registered, the agent automatically analyses its requirements and leaves a comment with a summary aligned with the company’s pre-sales standards.
Our new agent in action (use case):
I’ll walk you step-by-step through a pre-sales process in which our new agent actively participates and supports our team:
1. When the sales and technical teams meet with a client, all requirements and relevant information are documented on a page within the dedicated Confluence space. If there’s a Request for Proposal (RFP), it’s also attached to that same page.
2. After that, the Account Manager raises a pre-sales request in JSM through our internal portal so the technical team can prepare the proposal. In the form, they simply provide the URL of the Confluence page where all the requirements are gathered (Step 1).
3. Once the request is created, our Rovo Pre-Sales Assistant agent analyzes the pre-sales content and adds a comment with a summary aligned with the company’s standards. This is made possible by the automation we described earlier.
4. When our human agent handles the request, they already have a summary that speeds up their work. In addition, through Rovo Chat, they can interact again with our Pre-Sales Assistant to clarify questions or request support. For example, they can ask it to "suggest the project’s role distribution based on the profiles defined by the company"
5. Our team can not only ask the Pre-Sales Assistant questions, but also request that it generate a commercial proposal to serve as a base that can later be refined and sent to the client.
6. What’s even better, we can ask Rovo to publish the proposal directly in Confluence. This way, we ensure that all proposals remain centralized in one place, where they also serve as a knowledge base for future opportunities and, in turn, provide valuable feedback to the agent’s knowledge base.
7. From this point on, all that’s left is for our pre-sales team to review the document, enter the final pricing details, and send it to the client.
What have we achieved with everything we’ve seen so far?
I hope this use case inspires you to leverage Rovo within your company and imagine new scenarios where custom agents can empower your teams.
🌐 Spanish version in: https://www.linkedin.com/pulse/c%C3%B3mo-rovo-nos-ayuda-acelerar-el-proceso-de-preventa-1vszf
Nicolás Guzmán
0 comments