AI for real business problems: internal knowledge management 

One of the most common frustrations voiced by SMEs is the difficulty of finding the right information in their systems in a timely and efficient manner. Company policies, technical documentation, project guidelines and compliance materials are often stored in fragmented systems: SharePoint, Enterprise Resource Planning (ERP) tools, cloud drives, password managers, email chains. The result? Employees waste valuable time searching, asking around, or duplicating work. 

Several SME leaders in Jura are tackling this challenge head-on using AI-powered search tools. One company is building a secure, AI-driven interface that taps into internal data sources and responds to natural-language queries. Think of it as a private, company-specific ChatGPT that knows where your files live and what’s inside them. 

The system is built using open-source large language models (LLMs) like Llama, hosted entirely on local infrastructure to guarantee data security. Employees can ask, “Where can I find the latest version of the company travel policy?” or “What is the technical spec for client X’s network setup?” and receive accurate answers – sometimes even direct excerpts from relevant documents. 

One important lesson: the strength of this solution lies in the groundwork. It requires mapping, tagging and indexing company data, plus fine-tuning the model to prioritize the most current and reliable sources. While the AI can significantly reduce time spent searching, it won’t solve everything on its own – you need clean, well-structured data and clear boundaries for what the model can access. 

AI as a co-pilot: supporting – not replacing – employees 

A growing number of SMEs are shifting from viewing AI as a threat to seeing it as a collaborator. In day-to-day work, AI is proving especially useful in accelerating tasks that are repetitive, language-heavy, or code-related. 

SME owners, marketers and administrative teams are using AI tools to rewrite and translate texts, summarize meeting notes, draft client emails, and generate content ideas. Software developers report dramatic gains in efficiency, using tools like GitHub, Copilot or ChatGPT to write boilerplate code, troubleshoot bugs, or learn new frameworks. 

One SME founder described how an AI assistant helped him complete a complex JavaScript task in 20 minutes – something that would previously have taken half a day. Another company embedded AI directly into its ERP system to assist technicians: after completing a job on-site, the technician inputs brief field notes via a mobile app, and the AI polishes the text for clarity and professionalism before sending it to the client. 

Even basic tasks like preparing birthday messages or meeting summaries are being streamlined with generative AI tools. And with some experimentation, these tools are being fine-tuned to match a company’s tone, formatting style, or preferred language. 

The bottom line: AI doesn’t need to change your business model to create value. It can simply help your team do more, better and faster. 

That being said, people need to understand that a significant proportion – up to around a third – of responses from LLMs have been shown to be “hallucinations” or just plain wrong. You must always perform due diligence on the output of an AI before deploying it in the field or basing decisions on it. 

Building trust: data privacy and governance concerns 

Data security remains a top concern for SMEs, particularly those handling sensitive client or operational data. Many business owners share a strong reluctance to use cloud-based AI tools that lack transparency about where data is stored or how it is processed. 

In response, some SMEs are taking control by deploying AI tools on local servers or Swiss-based cloud providers. This approach ensures full control over internal data and aligns with Switzerland’s strict data protection standards. 

One IT integrator shared how they developed a proxy interface for ChatGPT-style tools. Employees access AI capabilities through a secure internal portal that enforces role-based access: HR staff see HR data, developers see code libraries, and marketers see campaign files. The interface looks and feels like ChatGPT but is deeply integrated into the company’s internal systems. 

Better still, this setup allows companies to shift from expensive monthly subscriptions to cost-effective pay-per-query pricing using APIs. It also allows management to audit usage and enforce security policies more easily, reducing the risk of accidental data leakage. 

As one participant put it: “It’s not just about having access to the tools. It’s about making sure they’re used wisely.” 

Facing the speed of change: how to keep up 

Another shared challenge is simply keeping pace with AI. With new tools, updates and models released almost weekly, it can be difficult for SME leaders to know where to start, what to invest in, or how to assess what is worth adopting. 

Several participants expressed a desire for more structured guidance – a map of the landscape. Which tools are best for scenarios? What are the trade-offs in cost, data control, and ease of use? How do you assess return on investment for a pilot project? 

This is where community exchange and peer learning can be particularly powerful. Sharing what worked, what didn’t, and how to take the first step can lower the barrier for others to begin. 

Best practices for SMEs getting started with AI 

Drawing on the experiences shared, here are a few best practices for SMEs looking to explore AI: 

• Start with a clear use case. Don’t try to “do AI.” Identify a pain point – a repetitive task, a bottleneck, a knowledge gap – and test a tool to address it. 

• Prioritize data security. If your business involves confidential data, consider open-source tools and local hosting options. 

• Involve leadership early. AI exploration needs a champion within the organization, empowered to make business decisions. In addition, if the boss isn’t curious about the potential of AI, the company is unlikely to move (fast). 

• Test before scaling. Try tools in a limited, low-risk context before expanding to broader use. 

• Encourage sharing. Set up regular team meetings or internal chat groups to trade AI tips and experiences. 

• Learn together. One of the strongest insights from the Jura discussions was the value of collaboration – many participants noted how sharing both successes and failures helped others clarify their own next steps. From informal discussions to structured workshops, exchanging real-world experience helps everyone move faster and more confidently. 

And most importantly: experiment. Every business is different, and the path to value is rarely linear. But the only way to learn what works for your company is to begin. 

Conclusion: AI is a tool, not a destination 

Artificial intelligence is not a magic wand, but it is a powerful toolkit. And in the hands of resourceful SME leaders, it is already making a difference. 

Whether by helping employees find information faster, speeding up content creation, or building smarter internal systems, AI can create real value when applied thoughtfully. The key is to stay curious, keep learning, and share what you discover. 

If you’re ready to start your AI journey, know that you’re not alone. Others in your region are already on the path, and their insights can help you avoid common pitfalls and accelerate your progress. 

Do you want to stay ahead of the curve? Sign up for our Industrial Transformation Innovation newsletter to receive invitations to future workshops, practical guides, and expert insights tailored for SME leaders navigating digital transformation.