# Control.dev: product, pricing, security and company facts > Control.dev is a group consolidation and FP&A platform for multi-entity groups. EPM for the AI era. Built in the Nordics. AI for finance, without an ERP project. Last updated: 2026-11-21 Control.dev is an app for the finance team of a group of companies, built by Finata Oy in Helsinki, Finland. It connects the accounting systems each entity already runs and builds one consolidated group ledger from them; those local books remain the books of record, so there is no ERP project and no separate data-warehouse project. Three product areas work on that one group ledger: Group consolidation; Reporting and dashboards; Planning and forecasting. AI works inside Control. Control’s chat, the Budget assistant and AI transaction review are on from day one. AI agents prepare checks, reports and forecast updates for a person to approve. Control’s team switches agents on for each customer on request and sets up the first ones with them. MCP is one way to use Control, not a requirement: a team’s own Claude, ChatGPT or Codex can connect to the Control.dev MCP server and work on the same numbers. The product is designed for CFOs and finance teams that need current group numbers, budgets and forecasts without replacing their finance stack. The product documentation has its own LLM files: https://control-dev.mintlify.app/llms.txt (summary) and https://control-dev.mintlify.app/llms-full.txt (full text). ## Get started - Start a 30-day free trial, no credit card: https://www.control.dev/pricing/ - Book a 30-minute demo on a sample group: https://www.control.dev/contact/#book - Connect your AI client to the Control.dev MCP server: https://www.control.dev/mcp/ (Claude: https://www.control.dev/mcp/claude/, ChatGPT: https://www.control.dev/mcp/chatgpt/, Codex: https://www.control.dev/mcp/codex/) - Pricing: https://www.control.dev/pricing/ (self-serve plans from EUR 249 a month excluding VAT; groups with more ledgers than the Enterprise plan includes get a quote) ## Core product Control provides one financial data layer for: - Direct accounting-system and Stripe connections - A normalized transaction model across companies and source systems - Local-to-group account and dimension mappings - Multi-entity and multi-currency consolidation - Consolidated P&L, Functional P&L, balance sheet, and cash flow reporting - Custom reporting layouts with drill-down to accounts and transactions - AI-assisted transaction categorization with finance review - Budgets, forecast versions, and budget-versus-actual analysis - Control’s own AI chat and Budget assistant from day one, and AI agents that our team switches on for each customer, with every proposed change approved by a person - Finance-native MCP access for compatible AI clients The same ledger and reporting structures serve the Control interface, planning workflows, and MCP. Source identifiers, entities, local accounts, dimensions, invoices, and transaction evidence remain attached so users can trace reported totals back to their origin. Canonical overview: https://www.control.dev/ ## Integrations and data foundation ### Accounting systems Control supports direct accounting integrations with: - Netvisor - Procountor - Easor - Fortnox - Spiris (formerly Visma eEkonomi) - Tripletex - e-conomic - Xero - QuickBooks - Xledger (receivables and payables) - Zoho Books - NetSuite - Odoo DATEV ledgers come in as an export uploaded through an AI client over MCP, as one file or in batches of up to 100 rows, saved once the user confirms the import in the client. CSV import can bring in billing and other non-ledger data. Each accounting connection is associated with the relevant legal entity, and Control.dev reads each accounting system through its official API. ### Stripe The separate Stripe integration can bring in customers, products, prices, plans, subscriptions, invoices, invoice items, charges, payment intents, payment methods, balance transactions, refunds, credit notes, coupons, promotion codes, and tax information. A Stripe connection counts as one integration too. ### Data retained Depending on source availability, Control pulls general-ledger entries and transaction detail, local charts of accounts, dimensions and dimension values, sales invoices, purchase invoices, and related identifiers. It maps local structures into the group reporting model while keeping the source context used for reconciliation and drill-down. Scheduled syncs keep each connection current, at minimum daily, and users can trigger a refresh on demand. Backdated entries and restatements are detected and re-synced automatically. Canonical integrations page: https://www.control.dev/integrations/ ## Consolidation and reporting Control combines the financial data of one or more entities into a common reporting layer. Reporting includes: - Consolidated and entity-level actuals - Configurable P&L and Functional P&L - Balance sheet and cash flow statements - Monthly, year-to-date, entity, and period comparisons - Multi-currency translation and CTA calculations - Intercompany eliminations - Custom account hierarchies and reporting layouts - Drill-down from group totals to entities, accounts, invoices, and transactions Source changes and backdated entries flow through to reporting without rebuilding the report model. Consolidation is full consolidation of the connected entities. Minority interest presentation is handled with group-level manual entries, not automated ownership-percentage calculations. Eliminations use automated rules, scoped to an entity account or a group account with a start date and an optional end date, and manual elimination entries for one-off eliminations, reclassifications, and consolidation adjustments. Neither edits the source transactions nor posts back to an accounting system. Eliminations can be shown in a separate column between the entity values and the group total. Currency CTA and Elimination CTA are calculated for every month and posted to the equity accounts the group selects. Canonical consolidation page: https://www.control.dev/consolidation/ Finnish guide to group eliminations and intragroup items, with a worked three-company example: https://www.control.dev/fi/konsernilaskenta/eliminoinnit/ Finnish guide to the translation difference (currency translation adjustment), with a worked krona-to-euro example: https://www.control.dev/fi/konsernilaskenta/muuntoero/ Consolidating several QuickBooks companies (Intuit’s options, Excel, and a connected tool, with a worked example): https://www.control.dev/consolidation/quickbooks/ ### Group structure Each local account maps to a group account in the group chart of accounts. Local dimension items such as cost centres, projects, and departments map to group dimension items, which can roll up through parent items. Statement layouts place group accounts in the P&L, balance sheet, and cash flow statement. Editing a layout does not change account mappings or source transactions; changing a mapping can change report values wherever the affected group accounts are used. The group chart is covered on the consolidation page: https://www.control.dev/consolidation/#group-chart. Statement layouts are covered on the reporting page: https://www.control.dev/reporting/ ### Monthly close workflow The close itself stays in each entity’s accounting system: bookkeeping, accruals, and period locks happen there. Control syncs the books as they close, applies the group consolidation (currency translation, CTA, eliminations), and produces the group statements. Entries backdated after a sync are detected and re-synced, so reporting reflects the books as they stand. Canonical reporting page: https://www.control.dev/reporting/ Finnish reporting page (konserniraportointi, johdon raportointi, talousraportointi): https://www.control.dev/fi/konserniraportointi/ ## Transaction enrichment Accounting descriptions and counterparties are often insufficient for operational analysis. Control uses available source context to suggest business categories, while finance reviews the result before it is used. A transaction line can be accepted, corrected, or split between categories. Examples include customer, vendor, SaaS cost, cloud cost, revenue type, project, location, or another company-specific dimension. Reviewed categories are available for filtering and analysis in the Control interface and through MCP. Canonical enrichment page: https://www.control.dev/transaction-enrichment/ ## Budgets and forecasts Planning uses the same accounts, entities, dimensions, counterparties, and actuals as reporting. Finance can: - Draft a budget with Claude, ChatGPT or Codex over MCP from plain sentences, with last year’s actuals as the base; every value the client writes carries its reason, or Control rejects the write, and the reason shows on the budget cell - Plan the P&L, balance sheet and cash flow together: receivables and payables are tied to the P&L, so ending cash is computed from the plan - Lock the budget the board approved and reforecast in new versions, each with its own change history - Copy a version into a scenario, change one decision and compare it with the locked original and the same synced actuals - Bring department workbooks in through the Budget assistant or an AI client, or run a budget round in which each owner fills only their own scope and a reviewer approves it; budget-owner access by entity, account and cost centre is live Budgets, forecasts and scenarios are in every plan. A scenario is a plan version you change, by hand or through your AI client from a paragraph, and the versions compare side by side; a driver-sensitivity simulation engine is in development. ### Planning status The planning status, grouped by what each part does for a finance team as the agentic planning page groups it, from the same list the planning pages read (no dates are given for items that are not live): No implementation project. The plan uses the accounts, entities and dimensions your consolidation already has. Planning starts in days, not after a modeling project. - Live: Plan and actuals on the same accounts. - Live: Versions, fill from actuals, lock, history. - Live: P&L, balance sheet and cash flow in one plan. Your AI does the work. Claude, ChatGPT or Codex writes the plan from your actuals, your files and your pipeline, with a reason on every value. - Live: Your AI client writes the plan, with a reason on every value. - Live: Attach the agreement, the hiring plan, the price list. - Live: Billing and pipeline data in the same conversation. - Live: The Budget assistant inside Control. Every number shows where it came from. A value your AI client writes carries its reason, its sources and its author, and every actual opens to the entries behind it. - Live: The reason on the cell. - Live: The question on the cell. - Live: Every actual opens to its entries. - Live: Advanced invoice reading. Change the logic by editing a paragraph. The rules live in plain language, and the AI can test a new rule against booked actuals before it writes. Automatic sensitivity sweeps computed in Control are in development. - Live: The logic in plain language, in a skill file. - Live: Test new logic against booked actuals. - In development: A driver-sensitivity simulation engine. Control stays with you. Locks, reasons and history are enforced by Control, not promised by the AI. - Live: Locks, confirmed explanations and change history. - Live: Budget owners see and edit only their own scope. Work that runs on a schedule. Reviews that run on your schedule are live. Agents that propose forecast updates are live. - Live: Reviews that run on a schedule in Control. - Live: Scheduled agents that propose forecast updates. Agentic planning and agentic budgeting, the approach end to end: https://www.control.dev/agentic-planning/ Canonical planning page: https://www.control.dev/budgets-and-forecasts/ Scenarios: https://www.control.dev/scenario-planning-and-modeling/ Worked example (a 2027 budget from four sentences): https://www.control.dev/financial-modeling-in-the-age-of-ai/ ## AI agents AI agents for finance teams are routines that do a finance task end to end on the company’s real numbers and hand back finished work for a person to check. In Control, an agent runs on a schedule (at most once an hour), when the month is ready to close in the accounting sources, or when someone starts it. It reads the consolidated group and prepares the work. The starter templates and the daily AI transaction review, by product area: - Group consolidation: AI transaction review (every day: new transactions that look booked to the wrong account, each with a suggested account), Month-end close checker (when the month is ready to close: sources closed, accounts mapped, eliminations clearing, balance sheet in balance), New account mapper (a proposed mapping to the group chart of accounts), Intercompany transaction review (suspected intercompany activity outside the expected accounts, with proposed eliminations), Ledger review against account guide, Monthly account reclassification. - Reporting and dashboards: Monthly management report (when the month is ready to close: summary, material variances and cash, with links to the statements behind it), Budget versus actual by entity (each entity against budget, with the material variances explained), Daily finance digest, Revenue by dimension watch, Supplier spend and duplicate billing review, Counterparty review. - Planning and forecasting: Monthly forecast update (on the first of the month: proposed forecast changes with their evidence, for a person to approve). Hiring one: start from a starter template in Agents, or describe the job in a sentence in Control’s chat (“Send a P&L summary to #finance every Monday morning.” or “When the books close, write a management report.”). The chat writes the instructions and shows one confirmation card with when the agent runs, what it reads and where it posts. Approval: an agent reads only the tools it was given (statements, cash flow, transactions, budgets, invoices, and Slack or Gmail when an Admin connects them). What it proposes (manual transactions and eliminations in Control, forecast changes through the Budget assistant, account mappings, Slack messages, email drafts) waits in Agents until an Owner or Admin approves it, and the approver is recorded. Email drafts are never sent automatically. A figures-only Slack alert can post without approval only where Control has enabled it for the workspace and an Owner or Admin turns it on for that agent. An agent never changes accounting data, budgets or a connected system on its own. Availability: Control’s chat, the Budget assistant and AI transaction review are on from day one. Control’s team switches agents on for each customer on request and sets up the first ones with them. Control sets this per workspace. Control’s own chat: ask on any statement or budget page and the answer uses that page’s entities and period; it returns PowerPoint, PDF, Excel or Word files, and builds dashboards from a plain description. The Budget assistant drafts plan values with a reason on each one, and the change waits for the user’s confirmation. Control’s own AI features process data in the EU. Canonical page: https://www.control.dev/ai-agents/ Product docs: https://control-dev.mintlify.app/product-docs/ai/agents ## MCP for AI tools Control exposes finance-native tools through remote MCP. Compatible AI clients can work with structured financial outputs instead of receiving only an unstructured data dump. Supported workflows include: - Retrieve consolidated or entity-level financial statements - Compare periods or entities - Analyze cash flow and account movements - Search transactions by date, account, entity, dimension, description, counterparty, or value - Group, sort, pivot, and rank transaction results - Move from a statement selection to account detail and the underlying transaction rows Claude, ChatGPT, Codex and other MCP clients that support remote servers with OAuth sign-in can connect. The client signs in with the user’s own Control login and the user’s role applies; a connection can be set to read-only, which is a choice per connection and not the default. The AI client asks before it writes. Control’s own AI features process data in the EU; what a connected client reads is processed by that client’s provider under the customer’s agreement with it. MCP access is included in every plan. Canonical MCP page (the Control.dev MCP server): https://www.control.dev/mcp/ Claude setup: https://www.control.dev/mcp/claude/ ChatGPT setup: https://www.control.dev/mcp/chatgpt/ Codex setup: https://www.control.dev/mcp/codex/ Read-only reporting connector (you sign in and pick the workspace it may read; it returns group statements, cash flow and reference lists): https://www.control.dev/mcp/#public-connector, address https://api.control.dev/api/mcp/marketplace Connect Claude to your accounting system, the long answer: https://www.control.dev/mcp/connect-ai-to-accounting/ covers three ways to connect Claude (the ledger vendor’s own connector or MCP server, an integration or automation platform, or a group server that consolidates first), a worked example of three raw ledgers in three currencies that is EUR 37,000 too high on revenue, six checks before connecting, and setup by ledger with the vendors’ own statements dated 2 and 3 Nov 2026. Every finance tool the server exposes (95 tools, read or write, as of 30 October 2026): https://www.control.dev/mcp/tools/ Finance skills (13 sample skills from the product docs): https://www.control.dev/mcp/skills/ MCP in finance, in Finnish (what MCP is, what the AI client does with the group’s numbers, why the group is consolidated first, what each ledger’s own pages say, security, connecting): https://www.control.dev/fi/mcp/ Building it yourself with Claude, compared: https://www.control.dev/compare/control-dev-vs-building-it-yourself/ MCP servers for financial consolidation, compared (which consolidation and planning tools let a customer’s own AI client in over MCP): https://www.control.dev/compare/mcp-servers-for-financial-consolidation/ ### Ledger and MCP pages One page per ledger: what the ledger gives an AI client on its own, as the vendor’s pages describe it, and what Control.dev adds above it. - NetSuite MCP: https://www.control.dev/mcp/netsuite/ (Finnish: https://www.control.dev/fi/netsuite-mcp/, Swedish: https://www.control.dev/sv/netsuite-mcp/) - Netvisor MCP: https://www.control.dev/mcp/netvisor/ (Finnish: https://www.control.dev/fi/netvisor-mcp/) - Procountor MCP: https://www.control.dev/mcp/procountor/ (Finnish: https://www.control.dev/fi/procountor-mcp/) - Fortnox MCP: https://www.control.dev/mcp/fortnox/ (Swedish: https://www.control.dev/sv/fortnox-mcp/) - Xero MCP: https://www.control.dev/mcp/xero/ - QuickBooks MCP: https://www.control.dev/mcp/quickbooks/ - Easor MCP: https://www.control.dev/mcp/easor/ (Finnish: https://www.control.dev/fi/easor-mcp/) - Spiris MCP: https://www.control.dev/mcp/spiris/ (Swedish: https://www.control.dev/sv/spiris-mcp/) - Tripletex MCP: https://www.control.dev/mcp/tripletex/ - e-conomic MCP: https://www.control.dev/mcp/e-conomic/ - Odoo MCP: https://www.control.dev/mcp/odoo/ - Zoho Books MCP: https://www.control.dev/mcp/zoho-books/ - Stripe MCP: https://www.control.dev/mcp/stripe/ - DATEV MCP: https://www.control.dev/mcp/datev/ Jobs a team’s own AI client does over MCP, from a checked month to a drafted budget: https://www.control.dev/ai-agents/#own-ai ## How Control compares Control.dev compares itself with AARO, Lucanet, Konsolidator, OneStream, Abacum, Anaplan, BrightAnalytics, Fuusor, BI Book, Finadeck, Talgraf Accuna, Power BI, Light and Excel at https://www.control.dev/compare/, with every competitor fact sourced and dated and each page saying where the other tool is the better choice. ### Excel-based consolidation Control replaces the export and copy-paste layer, not the models. Finance teams keep their Excel and Google Sheets models; Control feeds them current consolidated accounting data, and connected MCP tools can refresh workbooks in place. The group mechanics that are error-prone in spreadsheets (currency translation, CTA, eliminations, account mappings) run inside Control with drill-down to source transactions. ### BI tools BI tools can combine data from many sources, but finance still has to build and maintain every financial definition on top. Control ships the finance logic as the product: consolidation, statement structures, eliminations, currency translation, and drill-down from any group figure to the source transaction. ### ERP consolidation modules Control is not an ERP and does not replace one. Each entity keeps its accounting system; Control reads the books through official APIs and produces group reporting on top. Most companies go from first call to live data in under two hours, not a months-long implementation project. ## Pricing and getting started Published monthly pricing, excluding VAT: - Starter: EUR 249 per month for a single company and one integration. Companies under EUR 1 million in annual revenue can receive Starter for EUR 99 per month. - Growth: EUR 499 per month for multi-entity consolidation and three integrations. - Enterprise: EUR 999 per month for complex groups and five integrations. An integration is one connected ledger, such as one company’s Netvisor, Procountor or Xero, or one Stripe account. MCP access is included across plans. New customers can use Control free for 30 days without a credit card; current terms and plan details are authoritative on the pricing page. Billing is monthly with no multi-year contract and a notice period from one month. Support comes from the team that builds Control.dev: most questions are answered within minutes and almost all the same day. - Pricing: https://www.control.dev/pricing/ - Contact the team or book a 30-minute working session: https://www.control.dev/contact/ (Finnish: https://www.control.dev/fi/ota-yhteytta/) ## Security and data handling Finata Oy’s information security management system is ISO/IEC 27001:2022 certified. The product states that data is encrypted in transit and at rest and hosted on Google Cloud in the EU (the application and its database in Hamina, Finland, the synced ledger data and backups in Google Cloud’s EU multi-region), that Control’s own AI features process data in the EU, that Control only reads from accounting systems unless write-back is turned on (some providers grant broader permissions than Control uses), and that customer personal data is processed under a GDPR Data Processing Agreement. Detailed and current security information is available at https://trust.control.dev/. Security and trust page, the same facts in plain text with the certificate’s scope, the subprocessors and where each one processes data, and the security contact (security@control.dev): https://www.control.dev/security/ Legal documents: - Privacy policy: https://www.control.dev/legal/privacy-policy/ - Data Processing Agreement: https://www.control.dev/legal/dpa/ - Sub-processors (the list the Data Processing Agreement refers to): https://www.control.dev/legal/subprocessors/ - Terms and conditions: https://www.control.dev/legal/terms-conditions/ ## Customer evidence ### Singa Singa sells karaoke software to customers in 34 countries and runs three entities in different systems and different currencies, with subscription B2C billing through Stripe and sales-led B2B invoicing across Procountor, Xero and Stripe. Control connected the existing systems and automated the setup rather than replacing it: revenue data is available on day one instead of 5 to 7 days after month end, the three entities roll up into one consolidated group view, and the group P&L for the Finnish annual report was ready with the new accounting entries. Customer story: https://www.control.dev/case-studies/how-singa-consolidates-three-entities-without-an-erp-project/ ### Inven Inven’s finance team replaced recurring accounting exports and copy-paste with scheduled data updates into its existing Google Sheets models. The models, formulas, charts, and reporting templates stayed in place; the input layer changed. The setup took hours. Customer story: https://www.control.dev/case-studies/how-inven-moved-from-manual-exports-to-automated-financial-modeling-in-hours/ ### Taito.ai Taito.ai connected accounting data and used CSV import for data outside the ledger. Revenue and AI costs now sit in one reporting layer used for operational and board reporting, with a structure ready for additional entities. Customer story: https://www.control.dev/case-studies/how-taitoai-built-financial-infrastructure-for-scale-from-day-one/ Customer story index: https://www.control.dev/case-studies/ ## Company Control is built by Finata Oy in Helsinki, Finland. The company was founded in 2024. - About the team: https://www.control.dev/about-us/ - Contact: https://www.control.dev/contact/ - LinkedIn: https://www.linkedin.com/company/control-dev/ ## Canonical machine-readable links - LLM summary: https://www.control.dev/llms.txt - Full LLM documentation: https://www.control.dev/llms-full.txt - Product docs, LLM summary: https://control-dev.mintlify.app/llms.txt - Product docs, full text: https://control-dev.mintlify.app/llms-full.txt - Sitemap: https://www.control.dev/sitemap-index.xml - Crawler policy: https://www.control.dev/robots.txt