The impact of Generative AI on Flemish SMEs

From reducing administrative workloads to improving customer interactions, Generative AI, the technology behind tools like ChatGPT, offers countless new opportunities. Since 2023, we have been helping Flemish SMEs maximize the use of this technology through inspiration sessions, workshops, and concrete Proof of Concepts (PoCs) that immediately deliver added value. Our goal? To make Generative AI tangible and practical so that every organization can benefit from this technology. In this blog, we share insights, present some successful practical applications, and highlight the current challenges of Generative AI for SMEs. Curious about what this technology can mean for your company? Then be sure to read on!

Getting started with ChatGPT

For many companies, using ChatGPT is their first introduction to Generative AI. Both business leaders and employees eagerly use it to generate texts, draft emails, or quickly find answers to their questions. However, the usage often remains limited to the basics. The knowledge to deploy ChatGPT for more advanced applications is often still lacking, while the possibilities reach so much further. Some examples:

• Building private "GPTs"
• Searching and comparing documents such as PDFs and Excel files
• Generating all kinds of output formats such as tables, CSVs, or images
• Exploring other AI models, such as Google’s Gemini or Anthropic’s Claude
• Applying prompt engineering to obtain better answers
• Using ChatGPT as a search engine on the web

‍

Although ChatGPT is very versatile, we see that companies sometimes have incorrect expectations about what it can do in specific business contexts. The GPT models are trained on large amounts of general information, which makes them less suitable for answering company-specific questions. Fortunately, there are ways to use Generative AI for such applications. For example, you can use techniques like Retrieval Augmented Generation (RAG) to securely unlock company data and provide results that are better aligned with specific business contexts. And we can, of course, help with that!

Microsoft CoPilot as step toward advanced AI integration

Thanks to the introduction of Microsoft CoPilot within Office 365, many companies are taking their first steps toward Generative AI. Microsoft CoPilot offers accessible AI functionalities within familiar Office applications, but in practice, companies are often initially disappointed by the tool's possibilities and, in some cases, even revoke their licenses. This is often due to poor adoption by employees on the one hand, and an unsuitable underlying document structure and permission structure on the other. With proper content governance and correct implementation, however, companies can certainly use this tool to achieve efficiency gains.

Microsoft is investing heavily in the technology, and we expect CoPilot to become even more powerful and popular in the near future.

‍

At the moment, Microsoft CoPilot offers the following applications:

• SharePoint: creating, editing, and targeted searching of SharePoint sites and pages

• Teams: real-time summaries and action points from meetings, directly in the context of the conversation

• Word: support for understanding and summarizing documents and assistance in writing, editing, and creating content during work

• Excel: help to unlock insights, identify trends, and create professional data visualizations in a fraction of the time

• PowerPoint: designing and formatting slides and presentations based on natural language commands

• Outlook: assistance in drafting and improving emails

• OneNote: enhancing notes and helping to structure, understand, and retrieve information

AI use cases for Flemish SMEs

In general, SMEs don’t know much about AI. They know that AI could offer many opportunities, but where do you start, and how do you derive value from it?

We believe companies should at least search for inspiration and advice about the possibilities and challenges of this technology. The next step is often to experiment with tools such as ChatGPT or Microsoft CoPilot, and to define targeted use cases.

In the meantime, Studio Fledge has invested in developing an AI environment to quickly and affordably create custom prototypes. This way, we can immediately show how AI can provide added value for Flemish SMEs, practically and tailor-made to their specific context. Here are some examples:

Automation and efficiency

The most obvious AI applications for SMEs are often aimed at automation and efficiency gains. These are usually repetitive tasks that, thanks to AI, can be completed much faster and more easily. A few concrete, detailed use cases include:

Generating documents

Automatically generating documents such as quotes and reports is often a first step. Microsoft CoPilot can help with this, but it often fails to create more complex documents or strike the right "tone of voice" with simple prompts. That’s why we build workflows tailored to the client, such as a chatbot that creates documents based on custom templates and similar documents from the past. We generate the documents through a Q&A session with the user, company data, and so-called Large Language Models (LLMs, or intelligent language models).

Searching company documents

Another popular application is searching company documents, for example within SharePoint or as an extension of other document management platforms. With a good folder structure and the right metadata, search bots can answer questions based on company data or refer you to the correct documents.

Mail triage

Many organizations have a general mailbox (e.g., info@...) that receives numerous emails daily. Manually sorting these emails is a time-consuming process. Until now, emails could be filtered by sender or subject, but not yet by content. Generative AI can interpret emails and determine which department, such as sales, support, or billing, should handle them. This saves time for the company and ensures that every email is followed up promptly.

Transforming PDFs, Excel files, etc., into structured output

Non-standardized documents are sometimes harder to process. For several clients, we are developing a system that transforms and processes these types of documents (PDFs, Excel files, etc.) and converts them into a desired output format.

Evaluating tenders

Another interesting use case is the "tender evaluator" we developed for a client, which allows the AI tool to provide advice on the feasibility of a tender. The tool analyzes the company's context, activities, financial requirements, and awarding criteria and provides clear advice tailored to the client.

‍Automating intake processes

For a company offering rehabilitation aids, we developed a prototype to automate intake conversations using speech-to-text. AI converts speech into a structured format such as text and categorizes the information so that an internal ERP system can automatically process it. The tool can also formulate advice about next steps, process timelines, and more. Additionally, the application identifies missing data, improving the quality of registration. This not only benefits intake conversations but is also applicable in other domains, such as patient intake in healthcare, complaint processing in customer service, and work order registrations for laborers.

Innovative solutions

Generative AI opens the door to innovation in products and services. Many organizations are therefore experimenting with new applications to expand their offerings. The technology makes it possible to develop prototypes faster and easily integrate existing AI models as part of software development. Moreover, there are countless AI tools that assist in creating content, such as designs, texts, and videos.

For a client offering video content, we developed an application where users can easily search the videos through a chat interface. This allows them to quickly and easily find the right information in the videos. This approach shows how Generative AI can not only speed up processes but also improve the user experience.

Challenges and investments

The implementation of Generative AI requires the necessary investments. SMEs need the right infrastructure and knowledge to maximize its benefits. This means investing in hardware and software, as well as in employee education and training. After all, there is a learning curve associated with integrating AI solutions, but companies that invest time and resources will reap the rewards in the long term.

Step-by-step implementation as the key to success

Our experience shows that a step-by-step approach, where companies start with small experiments, is the most successful. By working with AI experts and developing prototypes, growing companies and SMEs discover the possibilities of AI step by step without taking significant risks. This also helps employees become familiar with the technology.

‍

Generative AI is here to stay, and more and more companies are realizing that. Yet we often find that organizations are not yet fully aware of its possibilities and pitfalls. The accessibility of tools such as ChatGPT and Microsoft CoPilot ensures rapid adoption, but those who are not well informed about the possibilities risk disappointment.

The good news is that AI is now very accessible and affordable, allowing companies to quickly and easily experiment and discover where the technology can make a difference for them. We are seeing more and more Flemish SMEs using Generative AI intelligently, and we expect this trend to continue to grow.

That’s why we recommend companies thoroughly inform themselves about AI and develop a clear adoption strategy that fits their business.

Would you like to learn more about a specific use case, see a demo, or get advice? Feel free to contact us; we’re happy to think along with you about the possibilities!