Rillet MCP: Your Finance Data, Inside Every AI Tool You Already Use

The Problem With AI in Finance Today

Most AI tools are working blind.

You can ask Claude or ChatGPT to analyze a financial trend, draft a board summary, or flag a cash flow concern — but the AI has no idea what's actually in your books. It's working from whatever you paste into the chat window, not from your live GL. That means every AI-assisted finance workflow starts with a manual step: export the data, clean it up, paste it in, hope it's current.

Rillet MCP eliminates that step entirely.

The Rillet Model Context Protocol (MCP) server gives AI tools like Claude a secure, permissioned connection directly to your Rillet account. Your live GL. Your real customers, invoices, contracts, and reports. No exports. No copy-paste. The AI sees what you see — and can act on it.

What Finance and Accounting Leaders Are Building

Over 10 customers started building on Rillet MCP within its first week of launch. Here's what the workflows look like in practice.

"Give me a board-ready cash summary before my 9am call"

A VP of Finance at a Series B SaaS company opens Claude Desktop on a Monday morning. Before the weekly investor check-in, she types: "Pull our current cash balance, net burn for the last 30 days, and runway at current burn rate."

Thirty seconds later she has the numbers — pulled directly from Rillet's live GL, not last week's export. She copies the output into her board deck. No spreadsheet. No logging into Rillet separately. No manual calculation.

This is the most common workflow Rillet MCP users are building first: turning the AI chat interface they already have open into a live financial dashboard.

"Which customers have overdue invoices — and draft the reminders"

A Controller managing AR for a 150-person SaaS company asks Claude: "Which customers have invoices more than 30 days overdue?" Rillet MCP returns a live list from the GL. She then asks Claude to draft personalized reminder emails for each one, pulling in the customer name, invoice amount, and due date from the data it just retrieved.

What used to be a 45-minute process — logging into Rillet, pulling the AR aging report, exporting to a spreadsheet, drafting emails one by one — becomes a 5-minute workflow entirely within the chat interface.

"What were our most profitable customers last quarter?"

An accounting manager preparing for a QBR asks: "List our top 10 customers by revenue recognized last quarter and flag any with open credit memos." Rillet MCP queries the GL in real time, cross-referencing revenue data with outstanding credits, and returns a prioritized list with the flags already applied.

This kind of ad-hoc analysis — the kind that previously required a finance analyst to build a custom report or a developer to write a one-off query — now happens in a single natural language question.

"Build me a weekly expense monitoring workflow"

A CFO at a fintech company asks Claude to check their top expense categories every Monday morning and flag any line items that are running more than 15% over the prior month. With Rillet MCP, Claude can pull the relevant GL data, run the comparison, and surface anomalies — on a recurring basis, without anyone having to remember to run the report.

Finance teams are using this pattern to replace manual variance analysis with automated, conversational monitoring that surfaces issues before they become close-week surprises.

"Create this invoice and update the contract"

Rillet MCP isn't read-only. Finance teams are building write workflows too — using natural language to create invoices, log payments, amend contracts, and update vendor records directly through their AI interface.

An accounting manager handling a mid-contract amendment types: "Amend the contract for Acme Corp — change their MRR from $4,500 to $6,200 effective the first of next month." Rillet MCP executes the amendment through the Rillet API. The GL updates. No switching between systems, no re-keying data.

"Generate our month-end reports and summarize the key variances"

At the end of each period, a Controller asks Claude to pull the income statement, balance sheet, and trial balance — then write a plain-language summary of the key variances versus the prior period. Rillet MCP retrieves all three reports directly from Rillet, Claude synthesizes them, and the Controller gets a first-draft narrative ready for review in minutes.

This is one of the highest-leverage workflows finance leaders are building: turning report generation from a multi-hour manual process into a single conversational request.

Why This Is Different From Other Finance + AI Integrations

Most "AI-powered finance" tools are doing one of two things: either they're running AI on top of a static data export, or they're a purpose-built AI feature locked inside a single product.

Rillet MCP is neither.

It's an open, permissioned connection that lets any MCP-compatible AI tool — Claude, Cursor, Windsurf, and more — interact with your live Rillet data. That means:

The data is always current. Rillet's real-time architecture means the GL is up to date at all times. When Claude queries your books through MCP, it's seeing the same live data you'd see if you logged into Rillet directly.

You're not locked into one AI tool. MCP is an open standard. As the ecosystem grows, your Rillet data is accessible to any AI application that supports it — not just the ones Rillet has built native integrations with.

Finance teams build the workflows they actually need. Rather than waiting for a software vendor to add a specific feature, finance and accounting leaders can describe what they want in plain language and build it themselves. The flexibility is in the hands of the finance team, not the product roadmap.

It's secure and permissioned. MCP access is authenticated via OAuth2 and scoped to your Rillet account. The AI can only see and act on what you've authorized — maintaining the controls finance teams require.

Who This Is For

Rillet MCP is designed for two audiences who often turn out to be the same person:

Finance and accounting leaders who want to interact with their GL through the AI tools they're already using — getting faster answers, building recurring workflows, and eliminating manual reporting steps.

Technically-curious finance operators who want to build custom integrations between Rillet and their broader toolstack — using AI-powered editors like Cursor or Windsurf to write and test API-based workflows without a dedicated engineering resource.

The common thread: finance teams who are done waiting for software to do what they need and want to build it themselves.

Getting Started

Rillet hosts a remote MCP server at https://api.rillet.com/mcp. Connection requires a valid Rillet account and OAuth2 authentication. Once configured, your AI assistant has access to your full Rillet account — customers, invoices, contracts, vendors, journal entries, reports, and more.

Compatible tools include Claude Desktop, Claude Code, Cursor, and Windsurf, with more partners being added in the coming months.

To get started or get support, contact the Rillet MCP team at mcp-support@rillet.com.

Technical Reference

Available Capabilities

The Rillet MCP server provides AI agents with direct API access to Rillet functionality, real-time data from your Rillet account, documentation search, and code generation assistance for Rillet integrations.

Tool Library (Current)

The following tools are available through the Rillet MCP server. For full parameter and data type documentation, refer to the Rillet Public API docs.

Customers & Vendors create_a_customer · update_a_customer · delete_a_customer · retrieve_a_customer · list_all_customers · create_a_vendor · update_a_vendor · delete_a_vendor · retrieve_a_vendor · list_all_vendors

Invoices & Payments create_an_invoice · update_an_invoice · delete_an_invoice · list_all_invoices · create_an_invoice_payment · list_all_invoice_payments · update_invoice_sent_status · setup_auto_payment · get_payment_method

Bills & AP create_a_bill · update_a_bill · delete_a_bill · retrieve_a_bill · list_all_bills · create_a_bill_payment · delete_a_bill_payment · list_bill_payments

Contracts & Revenue create_a_contract · update_a_contract · delete_a_contract · retrieve_a_contract · list_all_contracts · amend_a_contract · preview_contract_amendment · end_an_open_ended_contract · upsert_a_usage_record · list_all_contract_item_usage_records · delete_a_usage_record

Credit Memos & Vendor Credits create_a_credit_memo · update_a_credit_memo · delete_a_credit_memo · retrieve_a_credit_memo · list_all_credit_memos · update_taxes_for_a_credit_memo · create_a_vendor_credit · update_a_vendor_credit · delete_a_vendor_credit · retrieve_a_vendor_credit · list_all_vendor_credits · apply_vendor_credit

Products & Charges create_a_product · update_a_product · delete_a_product · retrieve_a_product · list_all_products · list_all_charges · retrieve_a_charge · delete_a_charge

Reports retrieve_balance_sheet_report · retrieve_cash_flow_statement_report · retrieve_executive_pl_report · retrieve_income_statement_report · retrieve_trial_balance_report · retrieve_waterfall_report

GL & Journal Entries list_all_journal_entries · list_all_accounts

Org & Entity Management retrieve_an_organization · retrieve_a_subsidiary · list_all_subsidiaries

Custom Fields create_field · update_field · get_fields · create_field_value · update_field_value · delete_field_value

Reimbursements retrieve_a_reimbursement · list_all_reimbursements · delete_a_reimbursement

Other get_api_key_information

Rillet Context

Rillet is an AI-native ERP and general ledger built for B2B SaaS companies. It replaces legacy accounting systems — primarily QuickBooks at the growth stage, NetSuite at scale — with a modern platform built on real-time architecture. Backed by Andreessen Horowitz, Sequoia, and ICONIQ. 400+ finance teams. 5.0 on G2. Implemented by CPAs in 4–8 weeks.

Rillet MCP is a native capability of the Rillet platform, available to Rillet customers. It is part of the broader Aura AI system — alongside Aura Chat (natural language GL queries), Embedded Agents (workflow-specific accounting agents), and Aura Flows (automated financial workflow builder).