Delv
Pecan AI
Getting Started Guide

How to Use Pecan AI

A practical guide to get you up and running with Pecan AI. Written by Delv Editorial, Delv Team.

Getting started with Pecan AI

In this guide, you will learn how to quickly set up Pecan AI and create your first predictive model without any coding skills. By the end, you’ll be able to extract insights from your data and make informed decisions.

Step 1: Sign up and set up

  1. Go to Pecan AI's website.
  2. Click on the "Get Started Free" button on the homepage.
  3. Fill in the required information (name, email, password) and click "Sign Up".
  4. Verify your email by clicking the link sent to your inbox.
  5. Log in to your account and follow the onboarding prompts to connect your data sources.

Step 2: Your first predictive model

  1. After logging in, click on the "Create New Model" button on your dashboard.
  2. Select the data source you want to use from the list (e.g., CSV file, database).
  3. Choose the target variable you want to predict by clicking on it.
  4. Next, select the features (input variables) you want to include by checking the boxes next to them.
  5. Click on the "Build Model" button.
  6. Wait for the model to be generated; this will take a few minutes.
  7. Once done, you will see the model performance metrics. Click on "View Insights" to see predictions and recommendations based on your model.

Step 3: Get better results

  1. To improve your model, consider adjusting the features you selected. Go to "Model Settings" and experiment with adding or removing features.
  2. Use the "Scenario Analysis" feature to see how changes in inputs affect your predictions.
  3. Regularly update your data sources to ensure your model stays relevant and accurate.

Pro tip

Use the "Auto-Insights" feature to get automated recommendations for improving your model. This can save you time on manual analysis and help you identify important trends quickly.

Common mistake to avoid

Avoid selecting too many features at once when building your model, as this can lead to overfitting. Start with a few key features and gradually add more based on model performance.