How to Build an AI Powered Sales Assistant on Scout

Today, I’ll guide you on how to use Scout to build an AI assistant for your sales team. This assistant can help search for prospects and prepare for sales calls. Gathering as much information as possible about a company is crucial for successful sales call.

First, create an app on Scout. If you don’t have a Scout account, you can create one at https://www.scoutos.com. After logging in and creating an app, you’ll be directed to a page like this. Here, we’ll construct our app.

In this App, we will use SERP calls (Search Engine Results Page call, a method used to retrieve data from search engine results) to gather various pieces of information about the prospect. First thing is to start by identifying the information you want to collect.

To do this, open the LLM block. Set your parameters as follows: use gpt-4-turbo for the model, set the temperature to 0, the maximum tokens to 600, and the Response Type to Text. After setting these parameters, we can start crafting our Prompt.

Below is a prompt that you can input into your App. It includes basic metrics about the prospect in question. Feel free to customize this to your preference. However, ensure that the information is easily accessible from a search page.

Info from the web: {{linkedin.output}}
Business model info: {{business_model.output}}
Info from crunchbase: {{crunchbase.output}}
Extra funding info: {{funding.output}}
Pricing: {{pricing.output}}
Today is: {{global.current_date}}
provide the following info for company {{inputs.input}} 
- **Company Name:** [Company Name]
- **Business Model:** [B2B/B2C], [Brief description]
- **Platforms:** [Mobile/Desktop/Both]
- **Age:** [Number of years]
- **Employees:** [Number]
- **Funding:** [$ Amount in millions]

- **Linkedin:** [url]
- **Crunchbase:** [url]

Now that we have all our search terms covered we will then add a Web block for each of the inputs we want searched. Click the “+” sign and create a Web block. Ensure this block is positioned above the output block, you may need to drag it. Let’s name this block Linkedin Scrape, you also may want to change the slug to something easier to read, for this one we’ll go with linkedin. For the search term we will use this syntax and the site you wish to search.

site:linkedin.com {{inputs.input}}

For metrics where we want to search the web and not a specific site, the search term will look like this:

{{inputs.input}} Business model

Once you have added your web blocks for each “variable” your App should look something like this:

Now its time to try it out! You can interact with your new App in two ways, you can simply type in the name of the company in the input box at the top top and hit “Run” or can open a chat window, to do so, look to the upper right hand corner of the screen.

Here we can enter the company that we want to search and in no time your new AI prospecting assistant will have a summary of the company you have a call with in 2 minutes.

Stay tuned for more on this App development. Next we will learn how to save the searched companies to a collection!

Follow and stay up to date with helpful and easy workflows to make with Scout!