​
Quick Start Tutorial

Welcome to your getting-started guide for Kolena’s Agents tool. This tutorial walks you through the steps to build your first Agent. In this example, we will create one that focuses on analyzing customer interactions.

If you’d like to follow along, the files used in the tutorial may be downloaded here.

​
What is an Agent?

A Agent is an autonomous, intelligent assistant that helps you automate or enhance your processes. Agents work on your behalf to analyze documents, extract structured data, and generate insights, freeing you to focus on strategic tasks.

Before you start creating an Agent, explore your organizational settings under your user profile (the menu in the top right).

Notable menu items include:

  • Manage Organization (for admin users): Manage your organization, users, and monthly plan.
  • Documentation: Extensive guides on Agent setup, user management, FAQs, and prompt styling.
  • Logout: Logs you out of the Kolena platform.

​
Step-by-Step: Create Your First Agent

Let’s walk through creating a Customer Support Agent to analyze customer interactions.

​
Creating the Agent

​
Option 1: Build a New Agent

  • Go to your main workspace
  • Click Create Agent
  • Define your Objective in your natural language Example: Analyze customer interactions and behaviors

​
Option 2: Select a Prebuilt Agent

  • Go to your main workspace
  • Click Create Agent
  • Choose from an existing prebuilt agent option, such as: Customer Support Triage

​
Upload an Example File

Start by uploading an example of customer interaction (email, support ticket, etc.)

  • Click Add New Support Ticket (sample file of whatever you’re working with)
  • Upload a sample email or document
  • Hit Submit to move to the extraction setup

​
Extractions

Extractions are the basic building blocks of an Agent. Think of each Extraction as an independent request to retrieve, compute, or synthesize some information from the uploaded document(s).

​
Creating Extractions

Adding an Extraction is as simple as asking (in natural language) what you’d want from the AI. Kolena will intelligently use the context it has to turn that request into a prompt that precisely guides the AI into extracting the relevant information from the uploaded documents.

Once you’ve provided a description of what you’re looking for, you can:

  • Let Kolena auto-detect the data type (Create)
  • Or preview the extraction, and manually adjust the instructions and data type (Preview)

Once created, Kolena will also save the Extraction to the Agent, so that future file uploads will automatically follow the same logic.

​
Customize Classification or Data Type

Each Extraction will also include an output data type, which Kolena will match the data to when the Agent runs.

Data Types Available:

  • Text
  • Number
  • True/False
  • Table
  • Classification (the Agent will pick the value out of the options you provide)

​
Examples of Extractions

Some useful extractions for our Customer Support Triage Agent:

  • Customer Name (Text)
  • Issue Summary (Text)
  • Next Steps for Resolution (Text or Classification)

​
Advanced Options

There are a few options for more advanced extractions:

  • Comparison Doc: Upload internal SOPs or guides for consistency checks
  • Data Analysis: Enable if using structured Excel files or needing calculations
  • Web Lookups: Pull public information as supporting context

​
Using Your Agent

​
Add More Files

Once your Agent is ready:

  • Drag and drop additional support tickets
  • Choose Import as Multiple Tickets (recommended for broad analysis)
  • Or Single Ticket (for deep dive into one customer’s issues)

Kolena will automatically extract all of the previously defined Extractions from the new files.

​
Manage Your Files

From the Agent Runs page (Support Tickets in this example), you can take further action to manage the files for your Agent:

  • Delete files: Select all or specific ones to remove
  • Download: Export as Excel
  • Filter: By customer name, issue type, etc.

​
Reporting & Collaboration

In addition to overall view provided by the Agent Runs page, you can also click into a specific run to:

  • View extractions in Editor View
  • Modify or delete extractions at any time
  • Use the Report View to:
    • Print summaries
    • Share live links
    • See reasoning and citations for extracted insights

​
You’re Ready!

You’ve now created a powerful Agent that can:

  • Automatically parse your support tickets
  • Extract key insights
  • Organize customer issues and trends
  • Provide structured, shareable reports

Refine your Agent over time by:

  • Connecting your Agent to myriad data sources or destinations
  • Adding more extractions
  • Uploading more examples
  • Iterating on your prompts

Happy automating with Kolena!