# 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](https://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**.

![Click on the Create Solution button](https://seclai.com/static/images/docs/getting-started-01.3229297880.png)

3. Make sure **Use AI assistant** is enabled (toggle should be on).
4. 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.
   ```

![The completed form with all fields filled in](https://seclai.com/static/images/docs/getting-started-02.1314266111.png)

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

6. 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.

![Review the proposed plan and click Accept & Execute](https://seclai.com/static/images/docs/getting-started-03.1738568020.png)

7. Once complete, the AI creates a pipeline with the components you need. Click into the solution to view and customize it.

![The pipeline created by the AI assistant — you'll need to add your data sources](https://seclai.com/static/images/docs/getting-started-04.680544144.png)

## 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. <a href="/static/files/taskflow-feedback-sample.2498214879.csv" download>
     Download the sample feedback file
   </a>
   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**.

![Click on the Sources box to open the content source](https://seclai.com/static/images/docs/getting-started-08.1928976219.png)

3. Click **Upload Files** and select the file you downloaded in step 1.

![The content source page where you'll upload your data](https://seclai.com/static/images/docs/getting-started-05.2426774238.png)

![Select your file and click Upload File](https://seclai.com/static/images/docs/getting-started-06.1784083137.png)

4. Once uploaded, the file will be processed automatically. When complete, you'll see the file with a **COMPLETED** status.

![Your feedback data is now ready for analysis](https://seclai.com/static/images/docs/getting-started-07.3262479735.png)

5. 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.

![Go back to the solution and click Agents](https://seclai.com/static/images/docs/getting-started-09.1535635972.png)

2. You should now see the agent the AI assistant created (the exact name may vary slightly). Click the **Run Agent** button.

![Your generated agent should appear in the Agents list, with an arrow highlighting Run Agent](https://seclai.com/static/images/docs/getting-started-10.89562356.png)

3. 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?
```

![Paste the sample question into Agent Input, then click Run Agent](https://seclai.com/static/images/docs/getting-started-11.1772254571.png)

4. Click **Run Agent**. The run should take a few minutes.

5. 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.

![Example output highlighting bug severity, frequency, and prioritization recommendations](https://seclai.com/static/images/docs/getting-started-12.1345485660.png)

6. You have successfully run your agent. Save the Agent ID from the page header, then refine your prompt and integrate with [SDKs](https://seclai.com/docs/api-introduction#sdks) or [webhooks](https://seclai.com/docs/api-introduction#webhooks).

## Next Steps

- [Agents](https://seclai.com/docs/agents) — Learn how to configure your agent for production use
- [Agent Traces](https://seclai.com/docs/agent-traces) — Inspect run history, step outputs, errors, and credit usage
- [Agent Triggers](https://seclai.com/docs/agent-triggers) — Run agents on schedules or events
- [Agent Evaluations](https://seclai.com/docs/agent-evaluations) — Monitor output quality and reliability
- [API Examples](https://seclai.com/docs/api-examples) — Copy-ready integration patterns
- [Troubleshooting](https://seclai.com/docs/troubleshooting) — Fix common setup and runtime issues