Great accounting advice starts when the spreadsheet is already done
An accounting consultant’s real value is knowing what the numbers mean for a specific business, in a specific situation, and what to do about it. That judgment takes years to build. But the workflows around it, pulling data, mapping accounts, formatting reports and writing narratives can consume most of the working day before any of that expertise gets applied. A budget request alone can take hours to prepare before the first real conversation starts.
SEVR Insights had a clear vision: give consultants a platform that handles the preparation, so their domain knowledge goes further, and give it now. Not in two years.
Together with Devies, they built the platform with a 100% AI-first development approach, running multiple agents in parallel across design, development, and testing to compress a build that would typically take a large team over a year into four months. The result was a system that connects to a firm’s existing accounting data through Fortnox or SIE file import and generates budgets, forecasts, liquidity analyses, and management reports in less than a minute.
Impact highlights
● 90% estimated time saving for generating standard financial reports
● Liquidity and receivables analysis available on demand through your digital CFO
● From zero to a product reaching 85,000 end users in a few months
In the age of AI, you can not sell hours anymore. What clients pay for is the knowledge, the judgment, the advice. We built this so the people who know the business can spend their time on exactly that.
The consultant's expertise becomes the product
The hardest problem was not making AI generate a budget. Spreadsheets have done that for years. The hard problem was making AI generate a budget that a consultant with deep domain knowledge could put their name on in front of a client, without hours of verification first. The answer is not to remove the human step, it is to make it the only step that matters. The consultant reviews, corrects, and delivers. Their judgment is what makes the output trustworthy. The platform just means they arrive at that moment with 90% of the work already done, in 30 seconds instead of an afternoon.
That only works if the numbers coming in are reliable. Devies built a custom query language that sits between the AI and the financial data, replacing open-ended database access with a controlled set of structured functions. Each function returns pre-aggregated, deterministic results scoped to the right client. The model cannot browse freely, construct its own queries, or surface data it was not explicitly asked for. What it can do is ask precise financial questions and get consistent, traceable answers to reason from.
What this changes is not just time. It changes what domain knowledge gets applied to. Every client conversation becomes one where the consultant’s expertise is the product, not the preparation work behind it.
CEO and period reports
Board material, period summaries, KPI overviews, and written recommendations can be drafted from live accounting data in less than a minute.
Ask AI, digital CFO
When a client describes a symptom, high revenue, positive profit, tight liquidity, the AI identifies the likely root cause and surfaces the data to support it. The consultant validates and delivers.
Liquidity in real time
Actual cash position with scenario handling for what comes next. Consultants can answer “can we afford this?” in the meeting rather than after it.
Budget and forecast
AI-built budgets and quarterly forecasts based on historical data, pipeline, and seasonality, generated in around 30 seconds. What previously required a dedicated controller is now a standard add-on.
Account mapping
Every company structures its accounting differently, and those differences matter across every report. AI-assisted mapping handles that complexity automatically, so the financial model stays consistent across every client.
Receivable analysis
Ageing breakdowns, top overdue customers, and concrete actions on payment terms and invoice frequency, surfaced automatically. A reactive conversation becomes a proactive one.
CEO and period reports
Board material, period summaries, KPI overviews, and written recommendations can be drafted from live accounting data in less than a minute.
Ask AI, digital CFO
When a client describes a symptom, high revenue, positive profit, tight liquidity, the AI identifies the likely root cause and surfaces the data to support it. The consultant validates and delivers.
Liquidity in real time
AI-built budgets and quarterly forecasts based on historical data, pipeline, and seasonality, generated in around 30 seconds. What previously required a dedicated controller is now a standard add-on.
Budget and forecast
Actual cash position with scenario handling for what comes next. Consultants can answer “can we afford this?” in the meeting rather than after it.
Account mapping
Every company structures its accounting differently, and those differences matter across every report. AI-assisted mapping handles that complexity automatically, so the financial model stays consistent across every client.
Receivable analysis
Ageing breakdowns, top overdue customers, and concrete actions on payment terms and invoice frequency, surfaced automatically. A reactive conversation becomes a proactive one.
How the team outbuilt the competition
Three developers and one designer. Four months and a fully functional financial platform that larger teams typically take over a year to ship. That is not a coincidence, it is what happens when an AI-first mindset is applied to the development process itself, not just the product being built. But the speed was only possible because of how closely Devies and SEVR worked together. SEVR brought deep knowledge of the accounting industry and a clear product vision. Devies brought the technical execution. Short daily syncs, weekly planning, and fast feedback loops meant decisions that would normally take weeks were made in days.
From the start, every part of the workflow was built around one question: Can an agent do this? Not as an experiment, but as a design principle.
Developers worked on multiple tickets simultaneously, with AI agents handling the groundwork while engineers focused on architecture, logic, and technical decisions.
The design system was built from atomic components connected to Figma through MCP. That meant the right component was always available when developers needed it – already documented, already consistent, and ready to use.
Quality assurance was also rethought. Instead of waiting for dedicated QA cycles, AI agents identified many issues earlier in the development process, reducing rework and speeding up delivery.
The team didn’t simply add AI tools to an existing workflow. They redesigned how work moved through the entire development process ...
The team didn’t simply add AI tools to an existing workflow. They redesigned how work moved through the entire development process, so people spent their time making decisions, while agents handled repetitive execution.
The same philosophy is built into the product itself: accountants spend their time applying professional judgment, while AI takes care of the preparation.
Accountec, one of the largest platforms in the Swedish accounting market, chose to promote SEVR Insights directly to their user base of 85.000 on launch day. That is the real proof point. You do not get that endorsement with something half-finished. You get it by shipping something the market is ready to trust and rely on.
