AI for Business: Beyond the Buzz

By: Bobby Stevens

Artificial intelligence (AI) is all the rage, and for good reason. But as a NetSuite developer, how can this translate to your company? You might not be building the AI yourself, but it helps to understand the basics.

How AI Models Learn (The Simple Explanation)

Think of an AI model like a student. It learns by looking at lots of past data. With your NetSuite info, it might try to predict future sales trends, identify potential high-value customers, or spot patterns that could mean equipment needs maintenance soon. The more data it has, the better it gets!

One Easy-to-Grasp AI Model: The Binary Classifier

This model is like a super-smart coin toss. It makes “yes” or “no” predictions. Examples:

  • Customer likely to buy again? (Yes/No)
  • Product likely to be popular? (Yes/No)
  • Equipment likely to need repair? (Yes/No)

It’s not 100% perfect, but it gives a probability. Like “80% chance of ‘yes’”. This gives your company a starting point for smarter decisions.

Key Point: No Need to Be a Math Genius

Modern tools make AI more accessible. You don’t need to be a deep statistics expert to get started and see if this is valuable to your company.

EAMPLE:

Teaching the AI to Spot Busy Days

Imagine you want the AI to predict if a day will be busy. But the AI doesn’t understand “busy” like we do. We need to translate it into something it can work with.

Step 1: Defining “Busy” in Computer Terms

Let’s say a busy day is when there are over 100 support notes in NetSuite. We need this data:

  • Data Collection: A list of dates and the number of support notes for each date.
  • The “Busy” Label: Add a new column. If a day has <100 notes, it’s “0” (not busy). If it has 100+, it’s “1” (busy).

Step 2: More Clues for the AI

Just dates and support notes might not be enough for the AI to learn the pattern. Let’s add more features:

  • Break Down the Date: Instead of one date column, have:
    • Month
    • Day of Week
    • Day of Month
  • Holidays Matter: Add two more columns:
    • Is Before Holiday? (Yes/No)
    • Is After Holiday? (Yes/No)

Picture This:

You’re making a little table for the AI:

MonthDay of WeekDay of MonthIs Before Holiday?Is After Holiday?Is a Busy Day?
JanTuesday15thNoNo0
DecFriday23rdYesNo1
… and so on

Why This Helps: The AI can now look for patterns like “Tuesdays in December before holidays tend to be busy.” This makes its predictions way better!

Scenario: It’s November 28th, a Monday. Your company wants to know if it’ll be a busy day for support.

Here’s how the AI model might think:

  1. Gather Features:
    • Month: November
    • Day of Week: Monday
    • Day of Month: 28th
    • Is Before Holiday? Yes (Thanksgiving is coming up)
    • Is After Holiday? No
  2. Compare to Past Data: The AI looks through its training data, searching for similar situations:
    • Mondays in November
    • Mondays just before a holiday
    • Other times with similar patterns
  3. Calculate Probability: Based on what it sees in the past, the AI calculates a probability:
    • Example: “There’s an 85% chance that today will be a busy day (label ‘1’)”
  4. Making the Call: Your company wouldn’t use this blindly. But, knowing there’s an 85% chance of a busy day helps them make decisions:
    • Maybe schedule extra support staff
    • Proactively reach out to customers to head off potential issues
    • Prepare for a higher volume of calls and emails

Important Notes:

  • No crystal ball: The AI is not perfect. There’s still a 15% chance it could be a slow day.
  • The more data, the better: The more information you feed the AI model over time, the smarter its predictions will become.

Need Help? Let’s Connect!

I’m a certified NetSuite developer dedicated to making NetSuite work seamlessly for businesses. If you have any NetSuite development requirements, I’d be delighted to assist! Please feel free to reach out.

← Back

Thank you for your response. ✨