1 of 14

Building a successful AI Solution for LN

Jihed Ayouni

Manager, Science

LN User Group

LN

1

2 of 14

2

Everyone is struggling to navigate the choppy waters of the AI hype? A lot of companies are doing reorg, trying to figure out what kind of tools they need.

Well, if you haven’t been living under the rocks for the past 18 months especially since chat GPT was introduced, we all know that has been the buzz, and it’s everywhere, newsroom, boardroom discussion

But let's be honest, sometimes the hype gets a little… intense. Yes, we all can agree that there is so much good stuff going on in the Al space, especially now with the new AI trends, I’m sure all of you have heard about Gen AI and large language models, some are picturing robots taking over the world, right?

But with all this talk, a crucial question remains, what’s AI?

If you watched , terminator, and Ex-Machina her these are just some of my favorite movies and these are examples of AI, at the end AI it’s is just another approach to automation and it’s here to make your profit higher.�

We going to take a closer look at AI but from Entreprise stand point,

AI can help us do things more effectively, this, this, that …. But how can actually help you in where you are. How AI can help me in my organization, in my company�

3 of 14

Agenda

01

02

AI Overview & Implementation Journey

Overview of Artificial Intelligence (AI) concepts and typical implementation journey

AI in Action for LN and Typical Architecture

Brief AI platform overview and demo of solution in LN along with architecture behind implementation

3

4 of 14

PREDICTING BUSINESS

OUTCOMES

ML learns from existing data to make predictions or decisions.

CREATING CONTENT

Generative AI uses what it has learned from a large data set to create new, similar kinds of data.

AI Terminology: ML vs. Generative AI

PREDICTING BUSINESS

OUTCOMES

CREATING CONTENT

4

5 of 14

ML vs. Generative AI

Use Case: Predictive Maintenance

for Manufacturing Equipment

Traditional PM:

Inefficient and costly
Unexpected downtime and lost production
Unnecessary maintenance tasks
Wasted time and resources

Machine Learning Solution:

Predicts equipment failures before they occur
Uses historical and real-time data from �machinery sensors
Identifies patterns or anomalies indicative of potential failures
Provides optimal service scheduling

Use Case: Generative Design �for Manufacturing

Traditional Product Design:

Time-consuming and costly
Trial and error method to meet specific �requirements
Must balance requirements with innovation

Generative Ai Solution:

Uses algorithms to generate a wide range of �design alternatives Based on set parameters �and constraints
Uses vast design space to identify the most efficient and innovative Designs solutions that human designers might not consider

5

6 of 14

AI in Action for LN and Typical Architecture�

Infor LN User Group

6

7 of 14

Reduce waste

Applications in industry

AI can enable data driven decision making across the enterprise

Streamline business operations

Manage supply challenges

Optimize manufacturing

Improve inventory management

Maximize customer experience

Leverage IOT to monitor and tune discrete and continuous production systems
Aid with workforce skill gaps

Raw material forecasting to plan manufacturing
Customer order forecasting for long term planning

Minimize spoilage by tracking for the right stock
Right goods at the right time and place

AI based intelligent matching to help reduce data access/entry issues
Financial forecasting and planning

Recipe optimization based on raw material and orders
Assist with experimentation on new products

Real time product recommendations
Pricing based on market conditions

Leverage IoT systems to improve operational tasks
Leverage rich history of your business to plan and operate your business efficiently
Upskill workforce with augmented abilities to improve labor productivity
Drive business decisions proactively and prescriptively

Building a Successful AI Solution for LN

7

When we think about AI application in any industry, we are thinking about all the different types of personas that drive in making the business successful. From the quality control manager who is looking at reducing scrap and improving efficiency so that they can improve the product line. To the supply chain manager looking at reducing the holding cost of surplus inventory and having enough stock to sell at the right competitor price ultimately to a happy customer. It's also important for an executive who's trying to make informed decisions about growth and talent

And the list goes on and on

but here at Infor, we don't just chase the latest AI trends, but rather our approach to innovation is what Infor calls AI/BI/PI:But obviously, it’s about reducing manual processes and paperwork. and how do you leverage data-driven decision-making to drive productivity, profitability, innovation, and prescriptive analytics to grow business and navigate the complex workforce?

�Today, let's forget the tech jargon and focus on the practical side of AI. We'll explore how machine learning and other tools can make a difference in your daily work by using your company's data to drive real, measurable progress.

8 of 14

High Level Architecture Overview

How is AI platform leveraged?

8

On-prem

or Cloud

Business App

Data Lake

MT Infor OS Digital Business Platform

Solution Templates

Augmented Intelligence Solutions

ION Desk

Trigger �workflow on a schedule

Data Orchestration

Data transformation into specs

Business action layer

User

Enterprise connector �for on-prem

Additional

Sources

MES Data / Aggregated IoT

Other third-party data

Retrieve real time output from endpoint

Realtime

Optional

Batch

Real Time

Analytics
Homepage & In-context widgets
Email Notifications
Mongoose Dashboards
Chatbots
ERPs
InforGo / Mobile
3rd party

Model zoo

Batch

8

9 of 14

AI Implementation Journey

Building a Successful AI Solution for LN

Identify your �problem

Exploratory data �analysis

Deploy �model

Maintain �value

Stage �observable

Feature �engineering

Model �tuning

Scope

Well defined
Establish definition of success and value proposition
Agree on project exit criteria

Solution

Validate data is suitable to solve business case
Iterate to build business explainable models
Track towards value

Deployment

Deliver solution to end users
Monitor and set up feedback for continuous model improvement

9

10 of 14

10

11 of 14

Birst Dashboard

AI / ML PLATFORM

Embed the AI / ML results into Infor Analytics

Predictive analytics on asset failure example

Analytics showing failure probability with additional graphs comparing prediction vs. actual
Ability to drilldown to the asset in the ERP itself for further information

Building a Successful AI Solution for LN

11

12 of 14

Decision making extensions

Extending architecture to create smart decision-making experience using LN drill back from the Asset Intelligence dashboard to create work order

Building a Successful AI Solution for LN

12

13 of 14

Solutions designed to make Artificial Intelligence accessible to the enterprise.

Infor platform technology / Enterprise artificial intelligence

Infor LN Connect – May 2024

13

14 of 14

Thank you

Infor LN User Group

Infor is a global leader in business cloud software specialized by industry

infor.com

14