Seclai

Seclai

Sign inStart for free

Getting Started

Getting Started

In this 5-minute tutorial, you'll build an AI agent that can analyze customer feedback, identify complaint trends, compare user segments, and surface actionable insights.

Prerequisites

You'll need a Seclai account. If you don't have one, sign up at seclai.com. If you already have an account, make sure you're logged in before continuing.

What You'll Build

By the end of this tutorial, you'll have an AI agent that can answer questions like:

  • "Compare the concerns of Enterprise users vs Free users"
  • "Which bugs should we prioritize fixing based on severity and frequency?"
  • "Write a summary for our product team about what we should address first"

The agent doesn't just retrieve text, it analyzes the feedback to generate insights you can act on.

Step 1: Create a Solution

Solutions group related agents, knowledge bases, and content sources together. We'll use Seclai's AI assistant to help set up everything you need.

  1. Navigate to Solutions in the left sidebar.
  2. Click Create Solution.
Solutions page with Create Solution button
Figure 1.Click on the Create Solution button

  1. Make sure Use AI assistant is enabled (toggle should be on).

  2. Enter the following:

    Name:

    TaskFlow Feedback Analyzer

    Description:

    Analyzes customer feedback to identify trends, pain points, and actionable insights

    AI assistant prompt:

    Create an agent that can analyze customer feedback for a SaaS product called TaskFlow.
The agent should be able to identify complaint trends, compare feedback across user
segments (Free, Pro, Enterprise), prioritize issues by severity, and provide
actionable recommendations for the product team.
Completed Create Solution form
Figure 2.The completed form with all fields filled in

  1. Click the send icon () to submit your request. It might take a few seconds to process.

  2. Review what the AI assistant came up with. If you want to change anything, you can add it in the text box below that section. Otherwise, click Accept & Execute and it will create the plan.

AI assistant proposing a plan
Figure 3.Review the proposed plan and click Accept & Execute
  1. Once complete, the AI creates a pipeline with the components you need. Click into the solution to view and customize it.
The created solution pipeline
Figure 4.The pipeline created by the AI assistant — you'll need to add your data sources

Step 2: Add a Content Source

The AI assistant has created your solution structure, including a content source named TaskFlow Customer Feedback (or a close variant, since AI-generated names can differ slightly). Next, let's add the sample data to that source.

For this tutorial, we'll use sample customer feedback from a fictional SaaS product called TaskFlow (a project management tool). The sample includes 50 pieces of feedback across different user tiers, categories, and sentiment levels. In a real scenario, you'd upload your own data or connect to RSS feeds and websites.

  1. Download the sample feedback file

     and save it somewhere easy to access, since we'll use it again in a moment.
  2. In the solution pipeline, click on the Sources box to navigate to the content source page for TaskFlow Customer Feedback.
Solution pipeline with arrow pointing to Sources box
Figure 5.Click on the Sources box to open the content source
  1. Click Upload Files and select the file you downloaded in step 1.
TaskFlow Customer Feedback content source page
Figure 6.The content source page where you'll upload your data
Upload Files modal with file selected
Figure 7.Select your file and click Upload File
  1. Once uploaded, the file will be processed automatically. When complete, you'll see the file with a COMPLETED status.
Content source with uploaded file showing COMPLETED status
Figure 8.Your feedback data is now ready for analysis
  1. Congratulations, you've just added your first content source! Your data is now indexed and ready for your agent to analyze. Let's put it to work.

Step 3: Run your Agent

Your agent is ready. Let's test it with some analytical questions that demonstrate its capabilities.

  1. Go back to the Solutions tab, then click the Agents option.
Solution page with Agents option highlighted
Figure 9.Go back to the solution and click Agents
  1. You should now see the agent the AI assistant created (the exact name may vary slightly). Click the Run Agent button.
Agents page showing the generated TaskFlow feedback agent
Figure 10.Your generated agent should appear in the Agents list, with an arrow highlighting Run Agent
  1. You should now see the Run Agent modal. This is where you'll enter the question you want the agent to answer. Paste this question into Agent Input:
Which bugs should we prioritize fixing based on severity and frequency?
Run Agent modal with example question entered
Figure 11.Paste the sample question into Agent Input, then click Run Agent

  1. Click Run Agent. The run should take a few minutes.

  2. When the run completes, you should see a response that summarizes the most severe and frequent bugs, calls out patterns, and gives clear prioritization recommendations.

Agent output showing bug severity and frequency analysis
Figure 12.Example output highlighting bug severity, frequency, and prioritization recommendations
  1. You have successfully run your agent. Save the Agent ID from the page header, then refine your prompt and integrate with SDKs or webhooks.

Next Steps