I am often asked: " Why do I have to deal with AI now? Are there really advantages with AI for my organization?"
Artificial intelligence has not really arrived in the business world yet. At least not in Germany. ChatGPT was the first to show the general public that AI has transcended from the slides of major consulting firms into reality and now plays a tangible role in everyday life.
Before the release of ChatGPT, the golden standard for large consultancies was to throw a few buzzwords on their slides. This often included topics like:
Machine Learning
AI
Industrie 4.0
Robotics
RPA
Blockchain
and much more
Buzzword-dropping by consultant led to AI as being perceived as yet another trend.
But artificial intelligence is here to stay. It's definitely not a buzzword, but likely the most revolutionary technology we'll see in the coming decades.
Where do I start with AI in my company?
The people who have recognized the urgency of AI transformation in their company quickly fall into despair or feel overwhelmed.
Many companies understand the urgency. But the implementation often causes panic. Management calls for 'AI, now!' Budgets are approved, and false experts appear. Yet, those responsible often feel left on their own.
When I'm asked, 'Where should I start?!' — I usually respond:
Artificial Intelligence has three main advantages:
It helps to automate manual tasks
It can be used as a personal assistant
It can create personalized customer experiences.
BUT...
To take advantage of any of these three benefits, you first need to start with the basics — processes and data.
Process Analysis and Assessment Steps for AI
Where do I start? I usually suggest starting by looking at the processes. The most immediate and noticeable benefit from using AI comes from reducing manual processes.
It's wise to analyze the core processes of each functional area. This information can later be evaluated by process experts to assess manual effort, digitization needs, and priority for achieving business goals. From this, a score can be calculated to determine the greatest benefit for implementing an AI solution.
What does this look like?
It's simple. You create a list of your company's core processes and start the evaluation.
Once the core processes are listed, the evaluation begins. Each process is closely examined:
How much manual effort is involved?
How digitized is the process already?
What role does this process play in achieving company goals?
These questions help determine a score that shows where AI can have the greatest impact - we call it the AI Impact Score. Processes that require a lot of manual work and are central to business success have the highest priority.
What next?
Then comes implementation. But step by step. There's no point in trying to digitize everything at once.
Instead, focus on quickly achievable projects, known as 'quick wins.' This shows early results, which also helps build internal acceptance for AI solutions in the organization.
Documenting business processes to be accelerated
After successfully prioritizing business processes based on the previously determined AI Impact score, start with a detailed process documentation for those with a high score (high score = high automation potential).
Only by thoroughly understanding the workflows can the right solution be implemented later.
Key questions to consider:
Who is responsible for the process?
What systems and tools are currently being used?
Which process steps are repetitive and could be automated by AI?
Where do bottlenecks or errors frequently occur?
Which process steps are critical to the business and should remain human-controlled?
The more detailed this process documentation, the more obvious AI opportunities will become. This allows for a targeted and effective AI implementation that leads to noticeable improvements in speed and efficiency.
Frameworks and Software for process documentation
To ensure a structured approach, process modeling tools such as BPMN (Business Process Model and Notation) or EPC (Event-Driven Process Chain) should be used. These models help visualize and analyze complex processes by mapping out individual steps and their dependencies.
For capturing and modeling, tools like Signavio (im Sap Kontext), ARIS, or Microsoft Visio. For process documentation in startups or SMEs we recommend Figma as it showcases a simple UI and good pricing model.
After the detailed process documentation is complete and clear potential areas have been identified, the next question arises: Which AI solutions are suitable for my company?
Choosing the right AI solution for my company
One thing must be clear: AI is not a universal solution, even though it is often perceived that way. I always say, AI is a building block — a part of an automation chain. Artificial intelligence can handle specific tasks very well, such as sentiment detection in texts, classification of inquiries (e.g., new customer inquiries, support requests), or processing large amounts of data. But these tasks are often just one part of the overall automation process.
One thing must be clear: AI is not a universal solution, even though it is often perceived that way. I always say, AI is a building block — a part of an automation chain.
Therefore, each use case must be considered individually. Implementing AI only makes sense when it is optimally integrated into existing processes and complemented by other tools and technologies.
The right approach often lies in the combination of different tools. AI should be used as part of an overall automation system, working hand in hand with other technologies.
A combination of no-code tools like Microsoft Power Automate, Zapier or Make.com can be a prcatical solution. These tools enable quick, flexible workflows that can be complemented by targeted AI deployments.
For specific needs that go beyond the capabilities of no-code tools, a custom-developed solution in Python — supported by its wide range of libraries — combined with AI solutions like ChatGPT can be a valuable addition. Python is ideal for such tasks due to its flexibility and wide application in data processing.
Pilot projects and test phases
After selecting the appropriate AI and automation solutions, it's wise to start with a pilot project. This involves testing a chosen solution in a limited area to ensure smooth integration.
It's important to use the combination of AI, automation tools, and programmable solutions in a targeted way. Only then can the full potential of the automation chain be realized, and the company's efficiency sustainably increased.
There is no single AI that automates everything. Many project managers from outside the field imagine it that way. If you take one thing from this article, let it be the understanding that AI is a building block in an automation chain, not a silver bullet. But it's a highly potent building block with unprecedented potential.
Get started with AI - we'll show you the way
Are you still overwhelmed with where to begin, what to prioritize, or simply need support?
At VISUS Advisory, we help you stay on track.
With targeted process analysis and customized AI solutions, we show you where the greatest automation potential lies and how to harness it quickly. Fewer manual tasks, more efficiency – and all in no time.
Don’t wait any longer. Contact us and kickstart your AI transformation with a clear plan and measurable results.
Comments