BigQuery MCP server: what an AI agent can query, and the guardrails it needs
Google's BigQuery MCP server gives an agent 9 tools, SQL on raw tables and up to 3,000 rows. It has no per-query byte cap and no metric definitions.
9 min read
#ai-data-access
31 articles
What an MCP gateway does (and what it can't fix)
An MCP gateway centralizes auth, call logging and rate limits for AI agents. It can't decide what a metric means. That lives in the semantic layer.
7 min read
Testing an MCP server before it touches production data (MCP Inspector and beyond)
MCP Inspector proves your server speaks the protocol, not that its answers are right. The rest: a scrubbed copy, an answer key, scope and cost checks.
9 min read
MCP vs API: what changes when an AI agent is the client
MCP doesn't replace your API; most MCP servers call one. What changes is who writes the call: a developer at build time, or a model at runtime.
6 min read
Snowflake MCP server: what the agent sees and what still needs a semantic layer
The Snowflake-managed MCP server exposes only the tools you declare. Answer quality still depends on a semantic view your team writes and maintains.
8 min read
15 best AI tools for data analysis for ecommerce in 2026
A ranking of 15 AI tools for data analysis on criteria C1 to C7. Ten of the 14 competitors need a data warehouse you have already built.
9 min read
14 best CDPs and alternatives for ecommerce in 2026
13 CDPs and Precisian scored on seven data criteria. None of the 13 documents versioned, customer-approved metrics. Activation is where CDPs win.
9 min read
Triple Whale alternatives: when Precisian fits, and when it doesn't
For a US Shopify brand, Triple Whale is usually the right call. Precisian serves e-commerce in Brazil, so it's the alternative only if you operate there.
8 min read
Triple Whale alternatives: 15 marketing attribution and analytics tools for 2026
15 Triple Whale alternatives ranked on seven criteria: data isolation, versioned metrics, MCP access, sources, who runs the data, and data ownership.
9 min read
How to give an AI agent customer context without exposing PII
Cohorts, segments and LTV all come from a pseudonymous ID. An agent almost never needs a name, email, phone number or address to answer them.
7 min read
Which AI agents can reach your data? Building the agent and MCP permission inventory
No console lists every AI agent touching your data. An MCP permission inventory needs seven columns and a monthly and quarterly review.
9 min read
What goes in a marketing data dictionary (and why your AI needs one)
A data dictionary records each metric's name, definition, source, owner, formula and exclusions. A 4 KB definitions file lifted AI accuracy 17 to 23 points.
7 min read
How to log every query an AI agent makes through MCP
Write the agent audit log on the server that runs the query. MCP logging was deprecated in the 2026-07-28 spec and was never built for audit.
7 min read
One AI agent, many client accounts: how to keep tenant data apart
When one agent serves many tenants, data rarely leaks through an attack. It leaks via leftover context, a cache keyed without the tenant, or one shared token.
8 min read
AI data governance: what the policy has to cover once an agent reads customer data
AI data governance decides what a model may read, not just who may use it. Five questions cover it: scope, isolation, retention, audit trail, metrics.
8 min read
ChatGPT and Claude for marketing data analysis: what they get right and what they make up
ChatGPT data analysis is safe for exploring a spreadsheet and unsafe for deciding with it: the math is right, the revenue definition is a guess.
8 min read
The official Google Ads MCP server: what it answers and what it can't see
Google's Ads MCP server reads your account through GAQL, read-only. It reports what Google logged as a conversion, not shipped orders, Amazon sales or returns.
7 min read
Google Analytics MCP: why the numbers it returns are already modeled
The Google Analytics MCP server reads GA4 through the Data API, which follows your reporting identity. On Blended, the numbers arrive already modeled.
8 min read
Why text-to-SQL fails on marketing data, even when the query runs
The query that runs is the worst case of text-to-SQL: the wrong revenue column, the wrong window or a fan-out join returns a clean number and no error.
9 min read
MCP security for marketing data: read-only, scopes, and the permission you shouldn't grant
Google's Ads MCP server is read-only, yet its OAuth scope can edit and delete your account. For ad data, MCP risk lives in the token's scope.
9 min read
Meta Ads MCP: the official connector, its permissions, and whose numbers it returns
Meta's official Ads MCP server entered open beta on April 29, 2026 with 91 tools. The conversions it reports are the ones Meta attributes to itself.
7 min read
Shopify's MCP servers are built for shopping, not for revenue questions
Shopify's official MCP servers act for buyers, customers and developers. None answers a merchant's margin-by-channel question; community servers fill that gap.
8 min read
Supermetrics alternatives when the consumer is an AI agent, not a spreadsheet
Windsor.ai, Funnel and Improvado all ship MCP servers now. Choose a Supermetrics alternative by who defines the metric, not by connector count.
7 min read
Connect Google Ads, Meta and GA4 to one AI and you get three revenue numbers
Google Ads counts purchases 30 days after a click; Meta, 7 days after a click or 1 after a view. Plug all three MCPs into Claude and one order counts twice.
8 min read
Teaching a data agent to say I don't know, and why that is hard
On a benchmark that prices wrong SQL, abstaining from everything scores 50% and the best real systems score 29.8% to 54.5%.
8 min read
Prompt injection through marketing data: the fields nobody validates
A Google Ads campaign name takes 256 characters with no documented content validation. Meta documents no maximum at all.
8 min read
Per-tenant isolation for MCP: giving access without giving away the warehouse
Row-level security fails at the role the agent connects with. Per-tenant isolation moves the boundary out of application code and into architecture.
8 min read
LGPD and AI: where your customer's data can live if you use a foreign LLM
To send customer data to an LLM outside Brazil there are two live legal paths, not three. Global corporate rules are not one of them.
6 min read
Query cost guardrails: what an AI agent can do to your warehouse bill
An AI agent can scan your whole warehouse without breaking a rule. Adding a LIMIT does not reduce what you are billed on non-clustered tables.
8 min read
What is a semantic layer? A field-by-field definition
A semantic layer is where your metric definitions live in a form a machine reads before it computes. Not a tool: a set of written decisions.
7 min read
Metric hallucination: why your AI reports a different number than your dashboard
Metric hallucination is AI computing correctly on a definition nobody wrote down. Supplying the definition raised accuracy from 45.5% to 67.7%.
7 min read