Data to AI™ / Gemini/Sage Accounting
Finance engineers and analysts pull Sage receivables, payables and journal totals into scripts, checks and notebooks from the terminal, with no Sage export on disk.
Better answers, fewer tokens.
Gemini CLI
Which customers still owe us for invoices that fell due in September?
Three customers still owe £18,450 on September invoices. Harlow Interiors accounts for about half of it, on an invoice due 4 Sep.
| Contact name | Total amount | Outstanding amount | Due date |
|---|---|---|---|
| Harlow Interiors Sales Invoices No Lines | £14,600 | £9,200 | 4 Sep |
| Brightwell Studio Sales Invoices No Lines | £7,900 | £5,350 | 12 Sep |
| Penrose Catering Sales Invoices No Lines | £4,100 | £3,900 | 19 Sep |
Read from Sales Invoices No Lines, one row per invoice, so totals are not repeated for every invoice line.
Sage Accounting, formerly Sage One, is cloud accounting software for small businesses. Dataddo extracts sales and purchase invoices, transactions, journals, ledger accounts, contacts and the product and service catalogues.
Query Sales Invoices No Lines for invoices with an Outstanding amount whose Due date passed in the last 7 days and write Contact name, Invoice Number, Due date and Outstanding amount to data/overdue.csv.
An operations analyst keeps a small script that drafts payment reminders. Each morning Gemini CLI refreshes the input file in the repo, and the analyst reviews the diff before the script runs.
Sum Outstanding amount per Contact name in Sales Invoices No Lines and Origin outstanding amount per Contact name in Transactions, then list the contacts where the two differ.
Before shipping a change to the receivables report, a data engineer asks Gemini CLI to compare two Sage datasets and pastes any mismatch into the merge request instead of deploying a wrong number.
Group Purchase Invoices with an Outstanding amount by week of Due date for the next eight weeks and add the table as a new cell in notebooks/cash-out.ipynb.
A finance analyst works in VS Code, where agent mode of Gemini Code Assist reads the same settings.json, and extends the cash planning notebook without leaving the editor.
Data to AI™ is priced per seat, like the AI assistants your team already uses, with ~25% off on annual billing. Need tailored contracts or volume pricing? Talk to us about Enterprise.
Serve governed business data to Claude, ChatGPT and other AI tools over MCP.
$10.49/month
per seat
3 seats minimum
Flexibly deploy Dataddo in any cloud or hybrid environment.
Custom
We have a payment model that works for you
| Without Dataddo |
With Dataddo
|
Outcome for you | |
|---|---|---|---|
| Input files for scripts | Someone downloads an invoice CSV from Sage, renames it by hand and drops it into the repo, so scripts break whenever a column moves. | Gemini CLI queries Sales Invoices No Lines by field name and writes the file with the same columns on every run. | Stable inputs, reviewable diffs |
| Sage access on laptops | Each developer who needs ledger figures asks for a Sage login or keeps a personal export in a downloads folder. | Only the Dataddo source holds the Sage authorization; Gemini CLI keeps a Dataddo sign-in in ~/.gemini/mcp-oauth-tokens.json and renews it automatically. | No Sage logins on dev machines |
| Control in the terminal | A script with API access runs whatever it was written to do, and nobody sees each request before it happens. | Gemini CLI asks you to confirm each Dataddo tool call, and you choose Always allow only for the tools you trust. | Every call visible |
| Config per repository | Connection details live in personal notes, so a new team member spends a day wiring up the reporting repo. | A .gemini/settings.json in the project holds the Dataddo httpUrl, so anyone who clones the repo signs in with their own Dataddo account. | Same setup for the team |
| Token consumption | Gemini spends tokens exploring raw Sage Accounting columns, sampling rows and retrying until it understands the data. | Field definitions and relations are supplied up front, so the model skips the exploration and goes straight to the answer. | Fewer tokens per answer |
| Answer quality | Prone to wrong joins, invented fields and numbers that are hard to check. | Answers grounded in defined fields. In the Dataddo benchmark, 78.5% of questions were answered correctly, against 65.5% with plain CSV files. | Better answers |
The Sage Accounting datasets finance and RevOps teams use most, with their real field names. Gemini queries them by name through the Dataddo semantic layer.
| Dataset | Metrics and attributes |
Ask Gemini
|
|---|---|---|
| Sales Invoices No Lines |
|
Which sales invoices with an Outstanding amount fell past their Due date in the last 7 days, as rows for a reminder script? |
| Sales Invoices |
|
Which Invoice line - ledger account values appear on sales invoice lines but are missing from our revenue mapping file? |
| Purchase Invoices |
|
What Outstanding amount on purchase invoices falls due in each of the next four weeks, by Due date? |
| Transactions |
|
How many transactions of each Transaction type were booked per day last week, so gaps stand out before a reconciliation job? |
| Journals |
|
Which journals have Journal lines - debit and Journal lines - credit totals that do not match, for a pre-close check? |
| Ledger Accounts |
|
Which Nominal Code values in Sage are missing from the account mapping file in this repo? |
| Contacts |
|
Which contacts have no Credit limit set, listed with their Credit days, for a credit-policy cleanup? |
We asked 43 pre-registered questions of three real datasets - Google Search Console, HubSpot deals and Google Ads - and delivered the same rows to the same model in four ways. Only the data layer changed.
of questions answered correctly - the best of the four delivery methods tested
fewer confidently wrong answers - 14.0% of runs against 24.0% and 26.3%
more correct answers on messy CRM data - better on 8 questions, worse on none
Same model, prompts and rows in every condition; 759 graded runs, ground truth frozen before any run. On Google Search Console alone, documented CSV files did as well as Data to AI™ - the advantage comes from messier data such as HubSpot deals.
Read the benchmark →Extract Sage Accounting in Dataddo with Sales Invoices No Lines, Purchase Invoices and Journals, then attach the flow to a Google Gemini destination or AI Model. Add the httpUrl https://headless.dataddo.com/mcp-data under mcpServers in ~/.gemini/settings.json, restart Gemini CLI and run /mcp auth dataddo to sign in.
In Gemini CLI, the url key falls back to SSE, which Dataddo does not support. Use httpUrl with no headers or oauth block, restart Gemini CLI, run /mcp auth dataddo and check /mcp.
Every flow or Context attached to an AI destination or AI Model in that Dataddo account, not only the Google Gemini ones. Keep Sage Accounting in its own account if other teams should not query it.
No. Sign-in uses your Dataddo email and password, with MFA if prompted. Gemini CLI opens the page in a browser, so run it on a machine that has one.
In VS Code, agent mode reads the same settings.json, so nothing changes. In IntelliJ, add the Dataddo server to the mcp.json of Gemini Code Assist instead.
Every Sage Accounting dataset re-reads a relative date window on each run. For older invoices in a backfill file, run a one-time full data re-sync with a wider range.
No. Gemini queries extracted data through the semantic layer by entity and field name. Dataddo never accepts SQL from the model and nothing is written back to Sage.
