“Why did our costs go up last month?”
Costs increased by €28,000. Cloud services and external contractors account for the change.
Next question: Show me the entries behind the cloud costs.
Example based on a monthly P&L comparison and its account movements. Actual answers depend on the available accounting data and your AI assistant.
For founders, CEOs and finance leads who need to understand what changed and what to look at next.
“Revenue grew. Why did our result get worse?”
Compare revenue, costs and the preliminary result between months. Identify the largest changes, then inspect the accounting entries behind them.
Go into your spending review knowing which costs to investigate.
“What happens to our result if revenue comes in 20% lower and costs stay the same?”
Use your reported revenue and costs as a starting point. Add your assumptions and ask your assistant to compare scenarios, showing the calculation and which figures come from your books.
See how sensitive your plan is before committing to the next step.
“Summarise last month’s financial performance and explain the biggest changes.”
Ask your assistant to draft a summary from revenue, costs and results, with the reporting periods and underlying figures. Add your commercial context and review it before sharing.
Spend your preparation time on the story and the decisions.
A founder’s next question
Start by asking which costs have grown and what your current cost base looks like. Use that history as the starting point for a hiring scenario in your AI assistant.
Then add what the books cannot tell you yet: today’s cash balance, the planned start dates and full cost of each hire, expected customer payments and other commitments. Ask your assistant to show its assumptions and compare the scenarios.
The useful output is a calculation you can challenge, with the missing inputs made explicit. Historical accounting data supplies part of that picture. Your plans and judgement complete it.
AI & MCP sits alongside the app, spreadsheets and API. Pick the workflow that fits the task.
| APX product | What you want to do | How you work |
|---|---|---|
| AI & MCP | Investigate a question and ask follow-ups | Talk to a compatible AI assistant connected to APX |
| Excel / Sheets Add-ons | Build and maintain your own financial model | Use financial data in a spreadsheet |
| Financial Data API | Integrate accounting data into your software | Build an integration with the REST API |
MCP is the connection that lets your AI assistant use APX data and tools. You keep working in your assistant; APX provides the financial data you are allowed to access.
APX MCP connection URL
https://app.apx.one/mcpWe’ll confirm the setup for your assistant and company when you explore the beta. Claude is the starting point. Access to other assistants, including ChatGPT, depends on their connection support and the current beta status.
Explore the beta with APX and confirm access for your company and chosen AI assistant.
Sign in at app.apx.one with the account you use for your company.
In your compatible assistant’s connector settings, add https://app.apx.one/mcp. The exact steps depend on the assistant.
Complete the APX sign-in and authorisation flow. The connection uses your APX permissions.
Start with a company and month. Compare revenue, costs and results, then ask which entries explain the largest changes.
A change in profitability, a hiring plan, a board meeting. Explore how APX MCP can help with your startup’s next financial question.
Build accounting data into your own applications.
Work with financial data in your spreadsheet models.
View your numbers and work with your APX team.
The bookkeeping behind your financial reporting.
Work through financial questions with the APX team.