In my previous post, I introduced the Generative AI capabilities available in Infor CRM v10 and some of the questions I wanted to explore as I spent more time with them. Since then, I’ve had a chance to get the integration configured and actually use some of these features.
Before getting into the details of configuring an AI provider, I thought it would be useful to first look at what these features actually add to Infor CRM.
One thing that became apparent pretty quickly is that Infor didn’t simply add an AI chatbot to the Web Client. There are several different places where AI has been incorporated into the application, each intended to help with a different part of day-to-day CRM work.
AI Summary
The AI Summary is probably the most immediately useful of the new features.
AI Summary is available for Accounts, Contacts, Campaigns, and Tickets. Rather than requiring a user to work through the various tabs and related records, it uses the information available in CRM to generate a summary of the record.
Depending on the record, this can include things such as activities, history, opportunities, products, and other related information.
For an Account, for example, the summary can provide a quick overview of what’s been happening with the customer without requiring someone to individually review recent activities, notes, opportunities, and tickets.
Account summaries can also include Engagement and Sentiment indicators. Engagement attempts to identify how actively the account is being worked, while Sentiment provides an indication of whether recent interactions appear positive, negative, mixed, or neutral.
The summary can also suggest actions based on the information it finds. Depending on the context, those actions can include scheduling an activity, drafting an email, creating an opportunity or ticket, or navigating directly to a related record.
What I like about this feature is that it isn’t trying to replace the underlying CRM information. It’s really providing another way to consume it. The source records are still there, but the AI can potentially save someone from having to piece the story together manually.
AI Assistant
The AI Assistant is closer to what most people probably picture when they think about adding AI to an application.
It provides a natural-language interface for working with CRM data. Instead of building a query or navigating through several groups and screens, a user can ask a question about the information in CRM.
The Assistant can return information, perform calculations, or produce lists of records. Depending on the results, it can also provide additional actions such as generating an email, scheduling an activity, starting a CRM workflow, saving the results as a group, or exporting the list.
That last part is particularly interesting to me. The Assistant isn’t necessarily just answering questions. In some cases, the result of the conversation can become the starting point for actually doing something in CRM.
There are obviously limits to what kinds of questions it can answer and how accurately it interprets them, and that’s something I want to spend more time testing. But even in its current form, it opens up some interesting possibilities for users who know what information they want but don’t necessarily know how to find it in CRM.
AI Newsfeed
The AI Newsfeed takes a somewhat different approach because it brings information from outside CRM into the picture.
It’s available on Accounts and can retrieve recent news and other information related to the company, then use AI to summarize those results.
Infor currently provides newsfeed options using GDELT, Google RSS, and Google Search. Administrators can choose which sources to enable.
This could be useful before a call or meeting with a customer. Instead of separately searching for recent news about the company, a user can potentially get a quick summary while already looking at the Account in CRM.
The usefulness will naturally depend on the company and the quality of the available results, but it’s an interesting example of AI being used to supplement CRM data rather than simply summarize what’s already stored there.
AI-Assisted Email Drafting
Generative AI also appears in the email process.
An email can be drafted from the AI Assistant, either by directly asking it to create one or by generating an email from a record returned in a list. AI Summary can also suggest drafting an email based on the context of the summary.
The generated message can use information already available in CRM to populate the recipient, subject, and body.
I don’t see this as replacing the user writing the email. The more practical use is probably getting past the blank page. If CRM already knows who the customer is and what has been happening with them, AI can produce a reasonable starting point that the user can review and edit before sending.
Built In, but Configurable
Another important distinction is that these features are built into Infor CRM, but the actual generative AI processing is provided by an external AI provider.
Infor CRM v10 currently supports:
- Google Gemini
- OpenAI
- Azure OpenAI
The administrator chooses the provider and supplies the appropriate API credentials. There are also configuration options controlling things such as token and request limits, the models used for different levels of AI processing, what CRM information can be included in summaries, and which external newsfeeds are enabled.
This means enabling AI isn’t quite as simple as checking an “AI Enabled” box, but it also gives organizations some control over how the integration is implemented.
I’ve been using Google Gemini for my initial testing, which has already given me a few useful lessons about model selection, API tiers, quotas, and how Infor divides requests between Simple, Moderate, and Complex AI functions.
That’s probably enough material for its own post.
Where I Am So Far
After spending some time with these features, I think the most interesting part is that there isn’t really a single “AI feature” in Infor CRM.
AI Summary is about quickly understanding a record. AI Assistant provides another way of finding and working with CRM data. AI Newsfeed brings external information into the Account. Email drafting uses the context CRM already has to help create communications.
Some of these will probably prove more useful than others depending on the organization and how people use CRM. I’m also still interested in seeing where the limitations show up once I start asking more complicated questions and testing against different kinds of CRM data.
For the next post, I’ll get more into the implementation side and walk through what I’ve learned configuring the Generative AI integration with Google Gemini, including a couple of things that weren’t immediately obvious when I first set it up.
