Request
Define the outcome
The operator describes the launch in business terms.
- Goal
- Purchases
- Market
- US and CA
- Budget
- $180 / day
AI agents for ads
Adrails gives paid media teams working agents, not a separate recommendation feed. They read the ad accounts and the business data around them, build a complete campaign, set up a split test or prepare the next budget and status changes from the figures.
Reviewed Oct 3, 2026
Product view
Move between the agent, cross-source analysis, campaigns and account context without rebuilding the brief in another tool.
Product workflow
The agent gathers the brief and account context, then turns it into the campaign structure the platform expects.
Request
The operator describes the launch in business terms.
Context
Saved rules and available assets complete the request.
Draft
The result becomes editable campaign entities, not prose.
Publish
The finished draft is created on Meta or Google Ads, paused until you switch it on.
Illustrative workspace data. Values show the product flow and are not customer results.
FAQ
Meta Ads and Google Ads, with the same four agents on both. Google Ads opens once Google finishes verifying Adrails. Connect Meta for now. TikTok Ads is coming soon.
It prepares campaigns and changes with their evidence, and you apply them. Automations can pause what crosses a threshold on their own, within the limits you set.
No. Adrails adds agents, automations and a cross-account workspace on top of each platform, which remains the advertising platform and system of record.
Yes. The Adrails MCP server connects Claude, ChatGPT, Claude Code and other remote MCP clients to your workspace. They read your accounts and prepare changes with the same tools and guardrails as the agents.