AI in mid-sized companies
Victor Wiedermann in conversation about the sensible use of artificial intelligence, process automation and how companies should start with AI in a structured way.
AI Enablement Programme
AI in your own business. Not a trial run, but a foundation to build on.
In six months we lay the groundwork for AI that answers from your own knowledge instead of generic training data. You finish with one use case already proven in production.
Free, no strings attached. Thirty minutes to work out where you actually stand.
What we build
Bring together
RAG systemAnswers from your own knowledge
Prepare knowledge
Knowledge baseProcesses, policies, experience
Create access
AI architectureSystems, interfaces, permissions
Each layer carries the next.
Companies that have already placed their trust in the founders of Scala Value
Process automation for over 20 years
Plenty of tools made it through the pilot. None made it into production.
The pattern is the same in almost every company we speak to:
Individual tools are tried out before anyone has looked at the processes behind them.
The AI answers in general terms because it does not know the company's own knowledge.
A pilot shines in the demo and falls over at the ERP interface.
Nobody can say what the effort achieved because nothing was measured.
The knowledge of experienced staff continues to be lost, tool or no tool.
Good AI output doesn't come from a good model. It comes from what knowledge the model can reach, and how cleanly it plugs into your systems.
That is where the programme starts, long before anyone talks about use cases.
Three steps, and none of them works without the one before
Together they form a foundation that further automation can run on, without starting from scratch each time.
Step 1
Create access
AI architectureWe look at your existing systems and data sources, the interfaces to CRM, ERP, ticketing, email and specialist systems, plus roles, permissions, data protection and operating model. The result works not only for a single use case, but also carries the next one.
Step 2
Prepare knowledge
Knowledge baseProcess descriptions, policies, product information, support knowledge, sales material, templates and the experience of long-serving staff: structured, checked and made usable for AI. Not a document archive, but the factual basis for everything that follows.
Step 3
Bring together
RAG systemThe connection between the two. The model no longer answers from general world knowledge, but draws on your checked content. Answers become more accurate, more consistent and, above all, traceable.
Why us and not the agency from last year
The market for AI automation is barely two years old. Plenty of providers can deliver a polished pitch. Far fewer can wire a system into an IT landscape that grew over decades without breaking it.
We come from IT, not from marketing
Our aim is to integrate AI. That requires understanding the existing enterprise architecture before designing an AI architecture on top of it.
Experience with critical infrastructure
Above a certain size, a tool running on its own no longer cuts it. We work in environments that not everyone is let into. That is where most providers run out of road.
Twenty years of process automation
Digitalisation and process automation were our business before anyone called it AI consulting. That kind of track record isn't something you can buy in.
This is where the AI products of Scala Value GmbH came from, among them SCALAHELP and SCALARECRUIT.
In the media
Sort out the goals before you shop for technology
Two trade publications have asked Victor Wiedermann how mid-sized companies should approach AI. Both conversations circle the same ground: spotting the potential that matters, the business case behind it, and how a first use case turns into something you can actually run. Both interviews are in German.
Victor WiedermannManaging Director of Scala Value GmbH
Article at wirtschaft tv
Using AI deliberately
Why companies should start with their processes and how artificial intelligence can be introduced in a structured way. Founders Magazin, issue 94.
Six months, six clear stages
Not a training course. A consulting and delivery engagement where we work alongside you, from the first idea to a prototype that runs.
A clear view of where you stand and where AI can actually help
What we do
Record the relevant processes and challenges
Analyse existing systems and data sources
Review available knowledge sources
Carry out an AI potential analysis
Assess the current level of automation
Prioritise use cases by benefit, effort and feasibility
What you walk away with
An overview of the most important AI potential
A prioritised list of use cases
A first assessment of effort and benefit
An overview of relevant knowledge sources
A shared target picture for the programme
What you have after six months
Not a deck about what might be possible. A basis for decisions, and technology that runs.
A clear view of the AI potential that matters
A prioritised AI roadmap
A target picture for your AI architecture
A structured knowledge base
A concept for your RAG system
One AI use case, worked out and tested
Documented processes, requirements and next steps
A foundation solid enough for whatever comes next
What gets built on a foundation like this
We built these on exactly that groundwork. They are not part of the programme and not what it promises. They show where the road leads once the architecture and the knowledge base are in place.
Standard queries about ticket purchases, refunds and access codes are handled end to end, wired into the ticket database and the event data. Anything complex is routed straight to the right person.
80%
of queries answered in real time, even during the rush before an event
Incoming CVs are matched against the job posting, suitable candidates are scored automatically and presented in the client's own corporate design, with the reasoning for why a profile fits.
90%
less time spent preparing candidate profiles
Every sales call was transcribed, speakers identified and the call script checked. That produced a score per agent and specific coaching recommendations, ready to hand over to Power BI.
1,000 calls
reviewed daily across roughly 30 agents, instead of manual spot checks
What the foundation pays off on
Once the knowledge base and the RAG system are in place, these stop being six-month projects and become a matter of weeks.
Internal AI assistant
Staff ask about processes, who owns what, policies or templates, and get the answer straight from your own knowledge base, instead of asking three colleagues.
Support assistant
Recurring customer queries are answered from checked product documentation, FAQs and service guidelines. Faster replies, consistent quality.
Sales and proposal assistant
Proposals, emails and arguments are drafted from your own service descriptions, references and pricing logic. Consistent, and without the manual writing.
Onboarding and offboarding
New joiners are guided through processes and tasks. At the other end, the knowledge of people leaving is captured properly instead of walking out of the door.
HR and recruiting
Role profiles, standard responses and internal rules sit in one structured place. Applications can be screened and candidate profiles drawn up.
Capturing knowledge before retirement
The specialist knowledge of long-serving staff is secured and made usable for others, before it leaves the building on someone's last day.
5,000 eurosper month, over six months
30,000 euros in total. The knowledge base and the RAG design are part of the deal, not an add-on.
What that looks like in proportion
One additional specialist costs you around 100,000 euros. Every year. The programme costs 30,000 euros. Once. And it leaves you with a foundation the next process can run on, without starting from scratch.
What you hold at the end is not a recommendation. It is an architecture, a knowledge base and a use case that has been tested. What goes into production is your call, made on the numbers from month five.
Common questions
No. It is a consulting and delivery engagement. We work on your real processes, systems and data, and build a working prototype with you. Your people pick up the knowledge along the way, because they are in the project.
Let's talk about where you stand
Thirty minutes, free of charge. Afterwards you'll know whether the knowledge base, the architecture or a specific process is the place to start. Whether or not we end up working together.
Pick a slot
Or drop us a line
We usually reply the same working day.