In the previous article we saw how the Frontier Firm represents a new organizational model, in which people and digital agents work together within business processes. But how does this transformation actually happen?
The cases presented by Microsoft show that the starting point is not necessarily a large Artificial Intelligence project. More often, it begins with a specific operational problem: processes that are too manual, information that is hard to find, AI tools that are barely used, or the need to build a stronger foundation of security and governance.
First simplify, then automate
In some projects carried out in the healthcare, financial and retail sectors, employees had to handle repetitive tasks every day across different tools: entering and transferring data between systems that weren't integrated, classifying requests, processing emails and manually updating information in management systems. The existing automations were spread across different platforms, hard to maintain and lacking common oversight. For this reason, the transformation started with technology consolidation, gradually moving automations and legacy tools to Microsoft Power Platform and introducing shared governance and analytics frameworks.
The results are concrete:
- migrating two automation platforms generated estimated savings of about $2.2 million a year;
- agents came to handle more than 50,000 emails a year, with over 9,000 working hours saved;
- process analysis identified roughly 40% of handled tickets as automatable.
Applying AI to fragmented processes risks automating inefficiencies that already exist. First you need to understand where to intervene, simplify the flow and build a coherent technology foundation.
Bringing AI into everyday work
A second case comes from retail, where a virtual assistant was developed directly inside the app already used by store staff. The goal was to help employees retrieve information on products and availability faster, support restocking and make onboarding new hires easier.
What's interesting is not so much the introduction of the assistant as the fact that it was integrated directly into the tools staff already use. Users don't have to open a new tool: they can access AI within their own workflow, using available company data while keeping traceability and control over operations.
For a fashion company, the principle transfers easily. A sales rep could ask the system about the availability of a specific size/color variant, a buyer could quickly check requirements and overdue orders, while customer service could retrieve information on orders and deliveries without navigating between multiple applications.
Training as the first step to truly adopting AI
A third case involves an enterprise company that had already made AI tools available but wanted to increase their actual use. An internal program was launched, built on training, daily use of prompts, sharing of use cases and AI champions tasked with helping colleagues experiment with new ways of working.
In nine months the program reached 91% adoption, with almost one million prompts in 28 days and about 25,000 examples of AI use shared internally.
The data shows that the transformation is not only technological. People need to learn when to delegate a task to AI, how to verify the result and, above all, where AI can produce a real advantage in their daily work.
What this means for fashion companies
Microsoft's cases show that there is no single path to the Frontier Firm. You can start by reviewing a process that is too manual, by introducing an assistant into the workflow, by running an internal adoption program, or by building a stronger foundation of security and governance.
What these experiences have in common is a very concrete approach: start from a business problem, use reliable data, introduce AI at the right point in the process and measure the result.
This is where Business Central and ModaSuite365 can play an important role for fashion companies. The ERP holds a significant part of the company's operational context (customers, suppliers, orders, availability, purchasing, costs and administrative processes), while ModaSuite365 adds the information and logic specific to the sector, such as models, colors, sizes, collections, seasonality, production and distribution.
This foundation can become the context in which to identify, step by step, the processes where AI can produce real value. It doesn't mean handing everything over to an agent, but understanding where people and digital agents can collaborate effectively and measurably.
The difference between experimenting with AI and becoming a Frontier Firm lies right here: bringing innovation into the company's real processes, one use case at a time.
Which processes in your company could be the first step? Contact us to find out how Business Central and ModaSuite365 can support a concrete path toward an increasingly agentic business model.
