Data to AI™ / Gemini/Salesforce
Give data engineers and RevOps developers Salesforce objects in the terminal, so pipeline snapshots, schema checks and duplicate audits land in the repo next to the code.
Better answers, fewer tokens.
Gemini CLI
Which open opportunities closing in September have not been touched for two weeks?
29 of the 85 open opportunities with a September CloseDate have not been modified in over 14 days. Qualification holds the most, with 17 stale deals.
| StageName | Opportunities | LastModifiedDate 14d+ | Stale share |
|---|---|---|---|
| Qualification Opportunities | 42 | 17 | 40% |
| Proposal Opportunities | 28 | 9 | 32% |
| Negotiation Opportunities | 15 | 3 | 20% |
Read from the replicated Opportunities object; stale means LastModifiedDate, the default replication date field, is over 14 days old.
Salesforce is a cloud CRM. Dataddo replicates the objects you choose, standard or custom, with the fields your org defines, or runs your own SOQL query.
Use the dataddo tools to describe Opportunities and tell me if every column listed in jobs/forecast/columns.yaml still exists there, custom ones included.
A data engineer runs this in Gemini CLI before merging a change to the nightly forecast job. A field renamed by the Salesforce admin shows up as a failed check in the terminal, not as a broken run at 3 a.m.
Query Opportunities with a CloseDate in the next 90 days, keep Id, Name, StageName and CloseDate, and write them to snapshots/pipeline.csv.
A RevOps developer saves one file each Monday in a small repo. Diffs between weekly commits show which deals moved stage or date, a history the live CRM view alone does not keep.
Find Accounts whose Name appears more than once and write each Id with its LastModifiedDate to reports/duplicate-accounts.md as a table.
Ahead of a merge project, an analyst working in VS Code with Gemini Code Assist hands the admin a reviewed list of duplicate records, newest edits first, inside the pull request.
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 | |
|---|---|---|---|
| Credentials on developer machines | Every script that talks to the Salesforce API needs a connected app, a client secret and a refresh token kept in a local .env file. | settings.json holds only the httpUrl. You sign in with /mcp auth dataddo, and Gemini CLI stores the session in ~/.gemini/mcp-oauth-tokens.json. | No Salesforce secrets in repos |
| Field lists that drift | Column names are hard-coded in scripts, and a custom field added or renamed by the admin breaks a job nobody is watching. | Gemini calls describe_entity on the replicated object, which reads its field list from your org, custom fields included, up to 300 per object. | Schema checked before it breaks |
| What the agent may run | A script with API access can call any endpoint it was written for, and nobody reviews the calls one by one. | Gemini CLI asks you to confirm each Dataddo tool call, and you choose Always allow only for the tools you trust, for example query. | Every call visible in the terminal |
| One setup per project | Each repo ships its own extraction code and API version, so two projects can report different pipeline totals. | A project-level .gemini/settings.json points at the same Dataddo server, so every repo reads the same replicated Salesforce objects. | Same numbers across repos |
| Token consumption | Gemini spends tokens exploring raw Salesforce 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 Salesforce 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
|
|---|---|---|
| Opportunities |
|
Which opportunities were modified today, listed with Id, Name, StageName and LastModifiedDate? |
| Opportunity Line Items |
|
How many opportunity line items were modified per day this week, based on SystemModstamp? |
| Accounts |
|
List the Id and Name of every account with a LastModifiedDate since 1 January. |
| Leads |
|
On which weekday are the most leads modified, based on LastModifiedDate? |
| Custom Objects |
|
Which Custom Fields are filled in on each of our Record Types, for a schema write-up? |
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 →Replicate Salesforce objects such as Opportunities and Accounts in Dataddo and attach the flow to a Google Gemini destination or AI Model. Add a dataddo entry with httpUrl https://headless.dataddo.com/mcp-data to ~/.gemini/settings.json, restart Gemini CLI, run /mcp auth dataddo and sign in with your Dataddo account.
In Gemini CLI, the url key falls back to SSE, which Dataddo does not support. Use httpUrl with https://headless.dataddo.com/mcp-data and add no headers or oauth block; /mcp auth dataddo handles sign-in.
Either. ~/.gemini/settings.json makes Dataddo available in every project, while .gemini/settings.json inside a repo limits it to that project, which suits a team repo that only works with Salesforce data.
No. Sign-in uses your Dataddo email and password, with MFA if prompted. Google and Microsoft SSO are not offered here, and Gemini CLI opens the sign-in page in a browser on the same machine.
Yes. In VS Code, agent mode reads the same settings.json as Gemini CLI. In IntelliJ, add the Dataddo server to the mcp.json of Gemini Code Assist, then ask about Opportunities or Accounts from the IDE.
Every flow or Context attached to an AI destination or AI Model in that Dataddo account, not only the Google Gemini ones. Keep flows with data a project should not see out of AI destinations.
No. Gemini asks for entities and fields by name through the semantic layer, Dataddo never accepts SQL from the model and nothing is written back. SOQL runs only in a Custom SOQL Query source you define in Dataddo.
Each run re-reads a relative window on a date field such as LastModifiedDate. Run a full data re-sync with a wider range once, and Gemini can query the older records too.
