What can companies use AI for?
Companies use AI mainly to process language, data and decision information faster. It is less about futuristic projects and more about daily processes that require too much manual work.
The three most practical tracks are these.
| Track | What it does | When useful |
|---|---|---|
| Hire Jan Kenis as AI consultant | Determining AI opportunities, choosing priorities and covering risks | When you don't yet know which AI case comes first |
| Build AI automations | Perform repetitive tasks automatically with AI | When team members often process the same information |
| AI software development | Integrate AI into your own tools, portals or workflows | When standard tools don't fit your process |
| what is GEO (Generative Engine Optimization) | Become visible in AI answers and AI search engines | When customers increasingly search via AI |
Which companies is AI suitable for?
AI is suitable for companies where there is a lot of knowledge work, customer questions, documents, data or marketing tasks. For an SME or growth company, the first step is usually not a large transformation program, but a well-defined application that saves time and can be monitored.
These situations often show that AI can deliver value.
| Situation | What AI can improve | Logic page |
|---|---|---|
| A lot of manual email traffic | Summarize, sort, prepare answers | Build AI automations |
| Spread knowledge and documents | Build an internal search function or knowledge assistant | AI software development |
| Teams use AI without agreements | Establish guidelines, prompts and quality control | AI training |
| Uncertainty about the best AI case | Choose use cases based on value, risk and feasibility | Hire Jan Kenis as AI consultant |
When do you start with an AI consultant?
You start by Hiring Jan Kenis as an AI consultant when you want to use AI, but do not yet know which application has the most value. A consultant maps out processes, data, risks and opportunities before you invest in tools, automation or software.
This is especially useful for SMEs where various teams are already experimenting with ChatGPT, Copilot or other AI tools, but where agreements regarding quality, privacy and measurable results are lacking.
Which processes do you automate first?
You first automate processes that occur frequently, have clear input and yield measurable time savings. Don't start with the most spectacular idea, but with the task that takes up hours every week.
Use these selection criteria.
- The task reoccurs daily or weekly
- The input is always similar, such as emails, forms or documents
- The result is controllable by a human
- Mistakes cost time, money or customer trust
- There is enough volume to recoup the automation
When do you choose custom AI software?
You choose custom AI software when a standard tool does not support your process well enough. This mainly happens when your own data, CRM, accounting, planning, customer portal or internal knowledge base need to be linked.
Customization is not an end in itself. Often the best solution is a combination of existing tools, APIs, and a thin layer of customization that does exactly what your team needs.
How does AI help with marketing and visibility?
AI helps marketing in two ways: work more efficiently and become more visible in AI responses. Efficiency involves research, briefs, content editing, segmentation and reporting. For visibility it is what is GEO (Generative Engine Optimization).
GEO builds on which is SEO. Pages should provide clear answers, explain entities, show resources and experience, and be technically readable by search engines and AI systems.
How do you practically start with AI?
You start practically with AI by choosing one process, measuring the current cost and then creating a small prototype. Only when that prototype saves time or increases quality do you scale up.
- Choose one process with clear input and output.
- Measure how much time and errors the current process costs.
- Determine which steps AI is allowed to take and which humans control.
- Build a small test with real examples.
- Compare time savings, quality and risk.
- Document the workflow before you scale up.
From practice
The companies that get the most out of AI start small and measurable. One task that takes twenty minutes every day will yield more annually than a large AI plan that never goes live. We therefore first look for silent time wasters: processing emails, qualifying requests, summarizing documents or transferring data.
Which risks do you need to cover in advance?
The main risks are incorrect output, privacy, dependence on external tools, unclear responsibility and poor data quality. AI should not blindly decide on matters where control, nuance or compliance is needed.
A safe AI project records these agreements.
- What data is and is not allowed to go to AI systems
- Which output always gets human control
- Who is responsible for errors or exceptions
- How prompts, workflows, and decision rules are documented
- How results are measured and improved
How do you choose between AI advice, automation, software and training?
You choose the right AI trajectory based on the question at hand. Advice is useful when uncertainty arises. Automation is useful for repeatable workflows. Software is useful when own data, rights and interfaces are required. Training is useful when your team mainly needs to learn to work better with existing AI tools.
Use this selection aid.
- Choose Hire Jan Kenis as AI consultant when you still have to determine the correct order
- Choose AI automation services when a process returns frequently
- Choose AI software development when customization, own data or a portal is needed
- Choose AI Training when your team needs to use AI securely and more consistently
Learn yourself or have AI built?
Learning yourself makes sense when your team wants to recognize and execute many small applications. AAI training then helps to see opportunities and work safely with tools. Having AI built makes more sense when connections, customization, reliability or scalability become important.
For a concrete process, we start with a free discovery call in which we focus on the process, the data, the value and the risks.