# Infor GenAI with Infor LN and M3: assistant and embedded features

> Source: https://netray.co/stack/infor-genai-assistant-infor-ln-m3
> Publisher: Netray (netray.co), AI and integration engineering for ERP and manufacturing systems.
> Section: ERP Stack

## Short answer

Infor GenAI is built on Amazon Bedrock. In LN it produces project executive summaries and contract views and answers assistant queries, in M3 it drafts service guidelines and product descriptions, and in CloudSuite Industrial it translates POs and invoices. Infor says the LN Assistant is in limited availability, and the service is cloud-hosted, which matters for on-prem and air-gapped plants.

## Erp

Infor LN

## Erp Key

infor-ln

## On Prem

true

## Product

Infor GenAI (GenAI Assistant and Embedded Experiences)

## Vendor

Infor

## Category

AI assistant and copilot

## Integration

**Method:** Embedded in CloudSuite through the Infor platform

**Detail:** These are Infor-delivered features inside the product, not a connector you build. They use Amazon Bedrock through Infor's platform, so availability depends on your Infor cloud deployment and entitlement.

**Method:** Assistant versus embedded experiences

**Detail:** The Assistant answers questions, while embedded experiences add specific actions such as summaries, descriptions and translations in context. LN Assistant availability is limited, so confirm what your tenant can use.

## Data Flows

**From:** Infor LN, M3 or CloudSuite Industrial

**To:** Infor GenAI service

**Data:** Prompt context such as project, contract, PO or invoice text

**Trigger:** User invokes the feature

**From:** Infor GenAI service

**To:** User

**Data:** Summary, draft or answer

**Trigger:** Response returned

**From:** User

**To:** ERP record

**Data:** Accepted text such as a product description

**Trigger:** User saves or approves

## Setup

### Confirm entitlement and availability

**Detail:** Ask Infor which features are available for your product, version and tenant, given the LN Assistant is in limited availability.

### Review the data handling terms

**Detail:** Understand what is sent to the service, where it is processed and whether it is retained.

### Classify your data

**Detail:** Decide which projects, contracts and items can be sent to a cloud model and which cannot.

### Pilot with a small user group

**Detail:** Choose a few users and use cases such as project summaries or item descriptions.

### Set review rules

**Detail:** Make a person approve any generated text before it is stored or sent to a customer.

### Measure usefulness

**Detail:** Track time saved and corrections needed, and drop features that do not earn their place.

## Pitfalls

- Assuming a feature seen in a demo is enabled for your version or tenant.
- Sending export-controlled or customer-confidential content to a cloud model without a policy decision.
- Generated item descriptions or translations saved without review, then reaching customers.
- Expecting answers across data outside the product scope, such as MES or PLM systems.
- Users treating summaries as authoritative without checking the source records.
- Features changing between releases so internal training goes stale.

## Ai Layer

### Private assistant for restricted data

**Description:** Where data cannot go to a cloud model, an open-weight model served on your own GPUs can answer questions over LN data, read-only, showing the query used.

### Cross-system questions

**Description:** A private layer can read LN together with documents and neighboring systems such as quality or maintenance data that vendor features do not cover.

### Draft-for-approval workflows

**Description:** Drafts descriptions, summaries and customer replies for a person to approve before anything is written to LN.

### Role-aware answers

**Description:** Answers respect LN roles so users only see data they could already see in the ERP.

## Alternatives

- Use Infor GenAI where your data policy allows cloud processing and the features cover your needs.
- Use plain LN reporting and BI for questions that are really structured queries.
- Use a private on-prem assistant where data residency or export control rules out Bedrock.

## Faq

### What does Infor GenAI do in LN?

**Answer:** Infor describes project executive summaries, contract views and answers to assistant queries in LN. The Assistant is described as limited availability, so check with Infor what your tenant and version can actually use.

### Does it run on-prem?

**Answer:** It is built on Amazon Bedrock through Infor's platform, which is a cloud service. Plants that run LN on-prem or air-gapped should ask Infor about availability and data handling before counting on it.

### Is it enough or do I need another AI layer?

**Answer:** If the embedded features cover your use cases and your data policy allows cloud processing, they may be enough. A private layer makes sense when data must stay inside your network or you need cross-system answers.

### Can it change ERP data?

**Answer:** The described features generate summaries, descriptions and translations. Any generated text should be reviewed by a person before it is saved, whichever tool produces it.

## Source

https://www.infor.com/blog/explore-the-latest-in-infor-genai

## Search terms

infor genai assistant ln m3, infor ln ai assistant, infor genai on-prem, infor m3 genai, infor ln copilot, infor cloudsuite industrial genai, infor amazon bedrock ai

---
Talk to an engineer: https://netray.co/contact?intent=erp_ai | Machine catalog: https://netray.co/agent/catalog.json