Integrations

Connect Adlyse to the tools your team already works in — so reports land in Slack, the assistant can read your CRM, and outside AI tools can work with your ad data. Here’s what plugs in and how.

Your ad platforms connect automatically#

When you onboard, Adlyse links your ad accounts using OAuth — the secure “Connect with…” flow that links an account without you ever handing over a password:

Google Ads, Meta Ads, and TikTok Ads — connected from Ad Accounts.

These aren’t add-ons you install. They’re the accounts Adlyse reads to power your dashboard, reports, and every recommendation.

Get reports where you work#

Slack — ask questions and route answers to the right organization and agent. Parent organization admins connect it from Settings → Slack.
Email — send reports to any address, and power teammate invites.

Let the assistant read your other tools#

One-click connectors give the assistant read access to another app you use, so it can pull in context alongside your ad data. Each shows up as a + Add tile under Integrations:

GitHub — browse repositories and read code, pull requests, issues, commits, and CI runs. Read-only, using your own GitHub permissions.
HubSpot — read your CRM records: contacts, companies, deals, and activities. (A CRM is your customer-relationship system.)
Airtable — read records and structure across the bases you can access.
Notion — read and edit the specific pages and databases you grant when you connect.
Stripe — read payments, customers, subscriptions, and invoices. Your live and test accounts connect separately.
Linear — read issues, projects, cycles, and comments from your workspace. Read-only.
Datadog — query monitors, dashboards, logs, metrics, and incidents. Pick your Datadog site (US, EU, …) when connecting.
Render — read services, deploys, logs, and databases. Connects with an API key from your Render dashboard instead of OAuth.
Google Sheets — read and write spreadsheet data across the sheets you can access.
SharePoint — read files and document libraries from your Microsoft 365 tenant.

Open Adlyse to outside AI tools#

There are two ways for AI tools outside Adlyse — Claude Code, Claude.ai, ChatGPT Desktop, or your own setup — to read and act on your data:

MCP servers — Adlyse’s set of Model Context Protocol (MCP) servers. MCP is the standard that lets outside AI assistants securely work with a product like Adlyse. This is the recommended path, and it has its own docs.
API keys — secret tokens that let those tools (and email delivery) authenticate back into your workspace. See API Keys.

Common flows#

Send weekly reports to Slack. Connect your workspace, then schedule delivery.
Give the assistant CRM context. Add HubSpot so it can read your deals alongside your ads.
Wire Adlyse into your own AI tools. Create an API key and point them at the MCP servers.

API keys

Create a token so outside tools can reach your data.

MCP servers

Connect Claude and other AI clients to Adlyse.

Reports

Set up the reports that go to Slack and email.

Your ad platforms connect automatically#

When you onboard, Adlyse links your ad accounts using OAuth — the secure “Connect with…” flow that links an account without you ever handing over a password:

Google Ads, Meta Ads, and TikTok Ads — connected from Ad Accounts.

These aren’t add-ons you install. They’re the accounts Adlyse reads to power your dashboard, reports, and every recommendation.

Let the assistant read your other tools#

One-click connectors give the assistant read access to another app you use, so it can pull in context alongside your ad data. Each shows up as a + Add tile under Integrations:

GitHub — browse repositories and read code, pull requests, issues, commits, and CI runs. Read-only, using your own GitHub permissions.

HubSpot — read your CRM records: contacts, companies, deals, and activities. (A CRM is your customer-relationship system.)

Airtable — read records and structure across the bases you can access.

Notion — read and edit the specific pages and databases you grant when you connect.

Stripe — read payments, customers, subscriptions, and invoices. Your live and test accounts connect separately.

Linear — read issues, projects, cycles, and comments from your workspace. Read-only.

Datadog — query monitors, dashboards, logs, metrics, and incidents. Pick your Datadog site (US, EU, …) when connecting.

Render — read services, deploys, logs, and databases. Connects with an API key from your Render dashboard instead of OAuth.

Google Sheets — read and write spreadsheet data across the sheets you can access.

SharePoint — read files and document libraries from your Microsoft 365 tenant.

Open Adlyse to outside AI tools#

There are two ways for AI tools outside Adlyse — Claude Code, Claude.ai, ChatGPT Desktop, or your own setup — to read and act on your data:

MCP servers — Adlyse’s set of Model Context Protocol (MCP) servers. MCP is the standard that lets outside AI assistants securely work with a product like Adlyse. This is the recommended path, and it has its own docs.

API keys — secret tokens that let those tools (and email delivery) authenticate back into your workspace. See API Keys.

Common flows#

Send weekly reports to Slack. Connect your workspace, then schedule delivery.

Give the assistant CRM context. Add HubSpot so it can read your deals alongside your ads.

Wire Adlyse into your own AI tools. Create an API key and point them at the MCP servers.