Every month-end, Dutch accountants running Yuki face the same ritual – exporting trial balances to Excel, copying invoice totals into reconciliation sheets, manually cross-checking VAT returns against bank feeds. It works, until you manage ten clients or fifty. This guide shows how to connect Yuki to a data warehouse and to Claude, so your month-end closes itself while you review the results.

What data lives in Yuki

Yuki stores the core financial data that drives every Dutch accounting practice. That includes purchase and sales invoices with line-item detail, journal entries across all daybooks, bank transactions matched to open items, VAT returns (ICP and OB declarations), contact records for debtors and creditors, the full chart of accounts, and scanned source documents linked to their bookings.

For accountants managing multi-administration setups, each client administration is a separate data silo. Getting a consolidated view – total revenue across all clients, average days-sales-outstanding by industry, or a list of administrations with unreconciled bank transactions – requires pulling data out of Yuki first.

Why getting data out of Yuki is harder than it looks

The practical result is that most accountants rely on monthly XAF or Excel exports pasted into spreadsheets. This works for a single administration, but it breaks down at scale. Every export is a snapshot that goes stale the moment new transactions arrive, and consolidating exports across clients means copy-pasting columns between workbooks – a reliable source of month-end errors.

Building a custom SOAP integration is technically possible, but the XML parsing overhead, rate-limit management, and per-administration session handling mean most small and mid-sized practices simply do not have the engineering resources to maintain it.

Since Visma Group acquired Yuki in 2020, the platform has focused on document recognition and bookkeeping automation rather than expanding its API or analytics capabilities. That is a sensible product decision, but it means external reporting and AI use cases depend on third-party integration.

The real cost of manual month-end

The bottleneck is not Yuki itself – it is the manual work that surrounds it. A 2024 BlackLine survey found that 30% of accounting teams still spend more than five days on month-end close. For practices managing 20-50 Yuki administrations, that translates to dozens of hours per month spent on export-paste-reconcile cycles that could be automated.

Manual exports also introduce error risk. Every time an accountant copies a trial balance into a spreadsheet, there is a chance of misaligned rows, stale data, or missed transactions. These errors are small individually, but they compound across administrations and often surface only during year-end audits – when they are most expensive to fix.

The data warehouse approach

The key difference from manual exports is that syncs run automatically. Once your Yuki connection is configured, Peliqan handles the SOAP calls, XML parsing, rate-limit management, and incremental updates behind the scenes. Your warehouse always has current data without anyone running an export.

For accountants managing multiple Yuki administrations, Peliqan syncs each one into the same warehouse. That means cross-client queries – total outstanding receivables, revenue per industry, average reconciliation time – become straightforward SQL joins instead of spreadsheet merges.

Once data is in the warehouse, you can transform and materialize tables to build clean reporting layers – for example, creating a unified client P&L view that combines journal entries, invoices, and bank transactions into a single table optimized for reporting.

From there, connect Power BI or any BI tool to the warehouse and build dashboards that refresh automatically. No more monthly exports – your reports update as Peliqan syncs new data from Yuki.

Skip the dashboard – query Yuki data with Claude

Not every month-end question needs a dashboard. With MCP (Model Context Protocol), you can ask Claude questions about your Yuki data in plain Dutch or English – and get answers in seconds rather than hours.

Peliqan’s Yuki MCP server connects your synced accounting data directly to AI assistants like Claude, ChatGPT, and Cursor. Install it with pip install mcp-server-peliqan and point your AI tool at the server – Peliqan handles authentication and data access.

The architecture matters here: Claude never connects to Yuki directly. Your Yuki data first syncs into Peliqan’s built-in warehouse, where it is governed, quality-checked, and combined with data from other sources. The MCP server then exposes this warehouse to Claude. This means the AI works with the same clean, consolidated data your dashboards use.

Once set up, you can ask questions like:

• “Which administrations still have unreconciled bank transactions this month?”
• “Show me total revenue per client for Q1 versus Q4 last year”
• “List all purchase invoices over EUR 5,000 that are past due across all administrations”
• “What is the effective VAT rate per administration for the current quarter?”

A practical month-end workflow might look like this: on the first business day, ask Claude to list all administrations with unreconciled bank transactions. Review the flagged items in Yuki, resolve them, then ask Claude to confirm the updated reconciliation status. Finally, ask for a consolidated trial balance summary across all clients – something that previously required hours of export-and-merge work.

For a deeper look at how Claude MCP works across different data sources, including setup instructions for Claude Desktop, Cursor, and Claude Code, see the full guide.

The MCP server also supports write operations. Through Peliqan’s reverse ETL, Claude can push data back into Yuki – for example, creating journal entries or updating contact records based on AI-generated recommendations. You can also build custom MCP endpoints for specific accounting workflows.

Comparison: four ways to get data out of Yuki

Get started with Yuki + Peliqan

Whether you want live dashboards, AI-powered month-end reviews, or both, the starting point is the same: connect Yuki to Peliqan’s Yuki connector and let the warehouse sync handle the SOAP complexity for you. From there, add Power BI for visual reporting, Claude for conversational queries, or both.

Peliqan offers fixed, transparent pricing starting at roughly EUR 199/month – covering the warehouse, connectors, transformations, and MCP server access. No per-seat charges, no query-volume fees.

Bonus: combine Yuki with Exact Online and AFAS in one view

Most Dutch accounting practices do not run on Yuki alone. Some clients use Exact Online for financial administration, others run AFAS for HR and payroll alongside Yuki for bookkeeping. Peliqan syncs all of these into the same warehouse, so you can build cross-tool dashboards and ask Claude questions that span multiple systems.

For example, you can combine Yuki and Power BI data in a single report that pulls live figures from all your client administrations. No more manual consolidation across Excel workbooks.

For AI-powered workflows, connect Exact Online through its own MCP server to give Claude access to both accounting platforms at once. Ask questions that span both systems – “Compare revenue recognition between our Yuki and Exact Online clients this quarter” – and get a single consolidated answer. With 250+ connectors and a 2-week custom connector SLA, Peliqan covers the full Dutch business software stack.