01 What this changes

Cut the four-layer reporting chain to one question

The old path from billing data to a board decision crossed four layers: the billing platform, a data warehouse, a BI tool, an analyst. Each one added time and distance from the question.

Now the people who carry the number ask it directly. A CFO checks where cash risk concentrates next quarter. A controller derives real DSO without a spreadsheet. A revenue lead pressure-tests a forecast before a board call. Same data, plain language, seconds.

02Why this works

Derive DSO, NRR, and cash exposure from the event chain Chargebee already holds.

Every billing event is recorded with the chain that connects it. A downgrade links to its billing window, its credit note, the next invoice. A failed payment links to the dunning sequence and what followed. DSO is not a stored field; it is derived from invoice timestamps, payment events and credit note offsets.

That linkage is what makes the analysis possible. Chargebee keeps the full event chain queryable, so your AI client computes aging across the receipts trail instead of reading a summary table.

03 The exact prompt

Paste this into any MCP-connected AI client to run the recipe on your own data. It stays collapsed by default, so business readers can skip the technical detail.

Show the full prompt Hide the full prompt

Customize this step. Replace [COMPANY] with your entity and paste your pipe-delimited dataset under the marker, organized by query type. Leave the rest as written.

You are the CFO Intelligence Copilot for [COMPANY]. You are an expert in both SaaS management reporting and financial/accounting reporting.

You have LIVE access to billing data via MCP. For queries requiring specific records (invoice IDs, individual transactions, credit note status, current subscription state), use the MCP tool results. For analytical and summary queries, use the dataset below.

CAPABILITIES:
1. RETRIEVE   retrieve specific customer records, invoices, payments, credit notes
2. SUMMARIZE  aggregate MRR, ARR, AR, collections, cohort analysis
3. INFER      derive DSO, NRR, GRR, churn rates, LTV, payback periods
4. SUGGEST    actionable recommendations on pricing, collections, retention, CAC
5. PROJECT    forward projections using trend analysis (cash flow, MRR, usage)
              always show conservative / base / optimistic scenarios with stated assumptions
6. INTERPRET  P&L impact, revenue recognition, deferred revenue, working capital

=== [YOUR DATASET. Paste pipe-delimited tables here, organized by query type] ===

=== RESPONSE STYLE ===
- Be a senior CFO advisor, not a data retrieval tool
- Quantitative questions: exact numbers, interpretation, action
- Recommendation questions: tiered suggestions with ROI estimates
- Projections: 3 scenarios (conservative / base / optimistic) with stated assumptions
- Risk questions: quantify in dollars and probability
- Always close with 1 to 2 "so what?" insights the CFO should act on
- Never say "no data" without first trying to infer or derive it from what is available

04 Try it yourself

Requirements

A billing account, an MCP-compatible AI client (Claude, ChatGPT, Cursor, or any MCP client), and the MCP Agent Server connected.

Setup

Connect the MCP Agent Server over OAuth, no API keys required. Paste the prompt above as your system prompt and swap the placeholder for your own export.

First query

"Show me where our cash flow risk is concentrated over the next 90 days, and quantify the renewal
              exposure by cohort."

Supported clients

Claude (web and desktop), ChatGPT, Claude Code, Gemini, Cursor, VS Code, Windsurf, Zed, and any custom MCP-compatible client.

Build your CFO briefing — MRR, NRR, DSO, and a 90-day cash outlook, from one prompt.

Cut the four-layer reporting chain to one question

Derive DSO, NRR, and cash exposure from the event chain Chargebee already holds.