Let’s start with a scene. Imagine a sales manager — we’ll call him Dieter — standing in a fluorescent-lit office in Düsseldorf, holding a lukewarm coffee. Dieter is staring at Salesforce. Salesforce is staring back, or rather, it’s spinning its loading wheel like a hypnotist’s pocket watch. Dieter has 10 reps on the phones, four in the field, and a director who wants yesterday’s KPIs yesterday.
The numbers are in there, somewhere, buried under drop-down menus, API connectors, “custom” dashboards that look suspiciously like the old dashboards, and of course, the monthly invoice that costs more than Dieter’s car lease.
The system isn’t broken. It’s doing what it was designed to do: sit between Dieter and the database, politely asking him to “click Save” before showing him a report. But Dieter doesn’t need to be polite. He needs the truth now, as a homeowner in Essen has just canceled a solar installation, and nobody knows why.
This is the moment — the “why is this so hard” moment — that explains the quiet revolution happening in enterprise software. Klarna has done it. A small German company called KreativReason.co is doing it in solar energy. And the revolution has a peculiar, almost sci-fi name: knowledge graphs.
SaaS was just CRUD with lipstick
For 20 years, the dominant model of enterprise software has been SaaS: Software as a Service. The idea was simple. Instead of installing SAP on your own creaky server, you rent Salesforce, HubSpot, HighLevel, or whatever else in the cloud.
Behind the curtain, though, SaaS is just CRUD — Create, Read, Update, Delete. Think of a database table: customers, deals, leads. SaaS builds a nice user interface, adds permissions, automates a few emails, and then charges you a recurring fee for the privilege of clicking “New Record.”
It was a good deal, until it wasn’t. As organizations grew, they bought SaaS for every slice of the business: marketing, sales, support, analytics, and HR. Suddenly, your company chart looked like a subscription graveyard. Each app had its own data silo, its own logic, its own password resets. The dream of “one source of truth” collapsed into a nightmare of Zapier connectors and quarterly integration projects.
Klarna’s graph epiphany
Enter Klarna. The Swedish fintech famous for “buy now, pay later” had its own Dieters — managers drowning in data spread across Salesforce, Zendesk, and a dozen other SaaS tools. Instead of buying more software, Klarna asked a radical question: What if the problem isn’t the lack of SaaS, but SaaS itself?
So they fired their entire SaaS team. Salesforce, gone. Zendesk, gone. The marketing stacks, the help desks, the endless dashboards — all replaced with a single layer: a knowledge graph.
Here’s how it works. A knowledge graph takes all of your company’s data — customers, payments, tickets, conversations — and connects it in a web, not a table. Instead of “rows in Salesforce” and “rows in Zendesk,” you have nodes and edges:
- Anna (customer) → bought → sofa
- Sofa → delivered → June 14
- Anna → filed complaint → damaged
Once the graph exists, you can ask it anything. Not with a clunky dashboard, but with natural language. Large Language Models (LLMs) sit on top of the graph like interpreters, turning questions into queries:
- Which products have the highest return rates this quarter?
- Show me all customers who called twice in the last week and still don’t have a resolution.
Klarna didn’t just save on SaaS licenses. They discovered they could run the company with fewer layers of abstraction, fewer integrations, and faster decisions. The database stopped hiding behind CRM screens and started talking directly to the people who needed it.
The eagle’s eye: Solar sales meet graphs
This isn’t just fintech magic. On the other side of Europe, in Germany, Hermann and Cynthia Rohr — co-founders of KreativReason.co — are applying the same model to an industry that is both ancient and strangely modern: solar installation.
The solar business is chaos wrapped in sunlight. You’ve got door-to-door reps knocking on doors in Bochum, call centers juggling inbound leads, field agents climbing roofs in Münster, and accountants in Cologne trying to reconcile who actually closed what. The data lives everywhere — Excel sheets, CRM records, WhatsApp chats — and managers often fly blind.
KreativReason looked at this and thought: What if we gave managers an eagle’s eye instead of a pigeon’s notebook? So they built a system on Neo4j, an open-source graph database. They wired in every event:
- Lead → booked appointment → door-to-door rep
- Appointment → converted → installation
- Installation → payout → finance
Then they added AI. Not “AI” as in “your chatbot will answer FAQs,” but AI as in “what if ChatGPT sat on top of your entire solar operation and explained it back to you?”
A manager can now ask:
- Which reps are booking appointments that never show?
- What’s the average time from lead to install, by region?
- Show me the projected revenue if conversion rates remain constant for the next quarter.
In other words, the database is no longer hidden under SaaS forms. It’s alive, connected, and speaking in plain German.
SaaS, meet your replacement
If SaaS was CRUD with lipstick, the new model is CRUD with wings. It doesn’t ask you to log in, click around, and export CSVs. It gives you a conversational interface to the truth.
The cultural shift is just as significant as the technical one. SaaS trained us to accept subscriptions as the cost of doing business. Graph + AI says: actually, your company already has the data. Stop renting it back from SaaS vendors.
This doesn’t mean every SaaS company is doomed. Many will survive as wrappers around specialized workflows. But the idea that Salesforce is the default nervous system of a company? That assumption is crumbling. Klarna proved it in fintech. KreativReason.io is proving it in solar. The pattern will repeat.
The hidden characters: APIs, Graphs, and LLMs
Here’s where we zoom into the invisible characters — the systems we don’t see, but live with every day.
- The API: For years, APIs were the diplomats between SaaS nations. They shook hands (or didn’t), brokered data swaps, and often failed at midnight. With a unified graph, the diplomats retire. Everything speaks the same language.
- The Graph: Think of it as the nervous system. Every neuron (customer, lead, sale) is connected, firing signals instantly. Unlike tables, graphs capture relationships, which is where business actually happens.
- The LLM: The translator, the friendly clerk at the counter. You ask for “top ten underperforming reps,” and the LLM converts it into Cypher queries or SQL. It doesn’t care about schema diagrams; it cares about answers.
Together, they form a stack that feels less like software and more like intelligence. Not artificial in the gimmicky sense, but artificial in the sense of augmenting human thought.
Why this matters
Step back, and this is bigger than solar or Klarna. It’s a story about how institutions remember things. For centuries, we built ledgers, filing cabinets, ERPs, and CRMs — all systems of record that required clerks (human or digital) to manage. SaaS was the latest software as a service (SaaS), available for rent at $99 per user per month.
But knowledge graphs with AI are more like memory itself. They don’t just store; they connect, infer, and explain. They move us closer to what organizations always wanted: not software, but situational awareness.
There’s humor in this, of course. We spent two decades building silos in the name of “integration,” and now we’re tearing them down in the name of “graphs.” Your IT director might roll his eyes and mutter: “Didn’t we try this with data warehouses in 2003?” Yes, we did. But warehouses were static. Graphs are relational in a way humans actually think.
The future: From CRUD to conversations
Picture Dieter again, back in Düsseldorf. But this time, instead of clicking through Salesforce, he asks: “Which of my door-to-door reps had three cancellations this week, and are they all in the same neighborhood?”
A voice replies — not Alexa, not Siri, but a domain-trained AI sitting on top of a graph. “Three cancellations. All in Essen. Same roof type, high shading. Recommend adjusting targeting criteria.”
That’s not a report. That’s a conversation with your business itself.
It’s more than just a report — it’s a conversation
And that’s why Klarna, KreativReason.co, and the next wave of companies are quietly stepping away from SaaS. Once you confront the truth directly, paying $150 per user per month to click through menus feels like paying rent on your own house.
KreativReason.co takes this logic and pushes it further. On the surface, they look like a small AI and full-stack development company tucked away in Halver, a German town best known for sauerkraut and a tendency to produce sturdy industrial products. But beneath that sauerkraut exterior, they’re building something far stranger: a new nervous system for business.
Instead of selling yet another CRM license, KreativReason builds microservices and full-stack applications that either integrate with your existing CRM or politely replace it. On top of this mosaic sits what the co-founders, Hermann and Cynthia Rohr, call the Brain.
If that sounds like science fiction, well, that’s the point. Imagine the “Main Ave” train station in The Matrix, where all the digital highways meet. That’s the Brain. Every system — CRM, call center logs, door-to-door sales apps, scheduling tools — feeds into it. Then the Brain makes it conversational. Suddenly, KPIs, forecasts, and financials aren’t static dashboards. They’re answers to questions anyone can ask.
Managers don’t click through tabs; they ask:
- “Which reps are wasting fuel driving across town when there are open leads nearby?”
- “How many calls did our AI bot make last week, and what percentage turned into real appointments?”
- “Show me which neighborhoods respond best to door-to-door offers, and adjust our targeting for tomorrow.”
It’s not software stacked on software. It’s more like a living network, harmonized by graph nodes, pulsing with company data like synapses firing in a brain.
And it powers things SaaS could never quite unify:
- A power dialer that doesn’t just call faster, but calls smarter.
- An AI voice bot that qualifies leads and books them directly.
- An intelligent scheduling system that routes sales reps by geography, not guesswork.
- A door-to-door app that transforms clipboards and spreadsheets into real-time intelligence.
Hermann Rohr puts it bluntly: “AI is a great leap forward, but it’s still just predicting the next token. That’s not enough for enterprise environments. The Brain changes this.”
His co-founder, Cynthia Rohr, expands the vision: “The future is API-first. And for AI systems to be knowledgeable, they need a knowledge graph — something that can actually learn from the business data.”
This philosophy guides everything KreativReason builds. They don’t just talk about data-driven decisions — they obsess over what they call data emission: ensuring that every interaction, every click, every customer call produces usable, connected, learnable information. Then they make that data accessible, not in the form of static reports, but in ways that actually help people run their businesses better.
In other words, KreativReason isn’t selling apps. They’re selling conversations — between humans and their data, between the past and the future, between the messy world of field reps and the clean logic of a graph database. And in that gap — between a lead in Dortmund and a KPI on a dashboard — they’ve built a little company in Halver that may just be shaping how the next decade of business software looks.
Closing reflection
In 1999, Salesforce ran ads saying “No Software.” It was cheeky, because of course, there was software — just in the cloud. In 2025, we might see companies running ads that say “No SaaS.” Not because there’s no software, but because software finally stopped being the bottleneck between humans and their own data.
And if you squint, you can see the irony. SaaS replaced the old servers by centralizing data in the cloud. Graph + AI replaces SaaS by decentralizing logic back into the organization itself. The pendulum swings, as it always does.
Meanwhile, Dieter sips his coffee. The cancellations in Essen are under control. His eagle’s eye is working. And for once, the database isn’t hiding. It’s talking.
Follow us on X, Facebook, or Pinterest
