Multi-agent · human-approved · language-agnostic

Marketing automation, sharp.

Automate & streamline your daily ad ops. A multi-agent harness plans the work, executes it across every account, and waits for your go-ahead on anything that writes.

No credit card · 2,500 free credits / month · invite your team in seconds
Approval-gated writes
Account memory
24/7 schedules
Multi-account context
01 · Multi-agent systems

A built-in multi-agent system, on a built-in AI Gateway

An orchestrator plus specialist agents — analytics, optimization, creative, research, safety and report — ship in the box. Each one runs on whatever model you choose, switchable any time on the Agents panel through one gateway.

  • Orchestrator + 6 specialist agents, pre-wired
  • Pick the model per agent — GPT-5.5, Claude, Gemini, Qwen, DeepSeek…
  • Built-in AI Gateway: hundreds of models behind one key
AI Gateway
GPT-5.5
Analytics
Optimization
Creative
Research
Safety
Report
Orchestrator
02 · Workflow orchestration

Every request becomes an observable, reusable workflow

The agent auto-clarifies the goal, judges what's needed, plans the steps, executes them, and emits a plan you can read. Plan → Execute → Summarize → Verify → Notify is generated for you — and when a change needs your call, it pauses and asks for approval.

  • Auto clarify · judge · plan · execute · output a typed plan
  • Plan → Execute → Summarize → Verify → Notify, automatically
  • Reusable workflows you can re-run, schedule, or share
Approval gate on every write
1Plan
2Execute
3Summarize
4Verify
5Notify
Auto-clarify
Observable
Reusable workflow
03 · Agent harness

A real harness — tools, skills, MCP, sandbox, memory, all automatic

Mid-run the agent picks the right Tool, Skill, MCP server and Sandbox on its own, and manages its own context and memory. It reads its execution history and the account, self-repairs errors from results, and keeps the task on track.

  • Auto-selects tools · skills · MCP · Python sandbox
  • Automatic context + memory management
  • Self-repairs from results; reads run history + account state
Tools
Skills
MCP
Sandbox
Memory
Context
Agent harness
auto-selects · self-repairs
04 · Schedule & run

Set it once. It runs 24/7.

Turn any task into a scheduled job — hourly, daily, weekly, or a raw cron in any timezone. Each fire re-evaluates against live data, so a recurring optimization never re-proposes yesterday's actions, and every run can email a summary.

  • Cron schedules with per-IANA timezone
  • Re-evaluates live data on every fire — closed-loop
  • Per-schedule approver email + failure alerts + run-now
24/7
cron · any TZ
MTWTFSS
Re-evaluates against live data each fire
05 · Teams & permissions

Organizations, roles, and everything scoped to your team

Create multiple organizations and set each member's permissions — admin, editor, or member. Manage your skills, integrations and assets per org, with cross-org sharing when you want it and isolation when you don't.

  • Multiple organizations · per-member roles (admin / editor / member)
  • Manage skills, integrations and assets per org
  • Cross-org sharing toggle · per-org keys, audit and billing
Organization
AAlex
Admin
SSam
Editor
RRiley
Member
Skills
Integrations
Assets
06 · Multi-account · multi-platform

One context, many accounts and platforms — executed together

The Context Picker lets you run across many accounts at once and pull Google Ads, GA4, BigQuery and Search Console into a single execution, with batch operations across them. More platforms are on the way.

  • Run across multiple accounts in one task — batch operations
  • Google Ads · Google Analytics · BigQuery · Search Console together
  • More to come: Meta, TikTok, and beyond
Context picker
Google Ads
GA4
BigQuery
Search Console
One run
batch · cross-account
+ Meta · TikTok
Why teams switch

Built on the opposite assumption from everything else

The problems it actually solves — and why teams don't switch back.

Reads are free. Writes wait for you.

Most ad AI either does nothing useful or quietly changes things you never sanctioned. Here every mutation pauses at a gate — in-app or magic-link email — and a single reject skips that step instead of aborting the run.

Decide on a source of truth you trust.

Google Ads conversion tracking is unreliable for real ROAS calls. Join Ads cost with your own BigQuery / CRM mart and let pause and budget decisions run off the numbers you believe.

It remembers your methodology.

A Skills library plus per-account memory — target CPA, protected campaigns, rules, history — captures how you operate, and the agent applies it on every run instead of starting from zero.

Move across accounts without breaking them.

One goal fans out across an entire MCC with a dryRun blast-radius preview first, v22 GAQL guardrails, sanity-checks, stale-state re-validation, and partial-failure tolerance.

Use cases

One agent. Many real workflows.

Concrete, multi-step jobs the harness runs end-to-end — each spanning several tools, accounts, and an approval gate.

Your workflow
Bring your workflow to optimize ads

The agent auto-reads GA4 and BigQuery (your CRM data) and executes the strategy from your custom skills — then routes any multi-account change to you for approval before it ships.

Guardrails
Hands-off account guardrails

Pause ads that consistently miss your CPA / CTR targets, with the plan and operations tailored from the account's history and its account memory — not a generic rulebook.

Audit & operate
Audit, optimize & operate

Auto-audit account performance, exclude irrelevant keywords, create RSAs and raise ad strength — then switch the bidding strategy once your goal is met, or spin up an experiment to tune bids.

Cross-platform
Cross-platform execution

Read your GA4 behavior data, build a matching GA Audience, and wire it into a remarketing campaign — the journey from analytics to live ads handled end-to-end.

Research
Research & strategy

Call the research agent to study the market and competitor sites, then craft a launch strategy — keywords, angles and budget — grounded in what it found.

FAQ

Questions, answered

Yes — security is our first priority. Every write pauses at an approval gate, every action lands in an operation / audit log, and you get email notifications for approvals and completions. Every integration, key and permission can be deleted and revoked at any time.

No. Reads are free; every mutation waits for a human — approve in-app, by magic-link email, or pre-approve a task when you explicitly want it hands-off. A single rejected step is skipped, not the whole run.

Yes. The orchestrator and each specialist agent run on whatever model you choose, switchable any time on the Agents panel through the built-in AI Gateway — GPT-5.5, Claude, Gemini, Qwen, DeepSeek and more.

Google Ads (deep, pinned to API v22), Google Analytics 4, BigQuery and Search Console today — with Meta, TikTok and more on the roadmap. Plus thousands of SaaS actions via Composio MCP and your own MCP servers.

Yes. Every workspace is an organization with isolated accounts, integrations, skills, assets, audit log and billing. Create as many orgs as you need, set per-member roles, and fan one goal out across an entire MCC.

Whatever you write in. Plans, reports and the agent's own memory all match your language — describe a goal in English, 中文, 日本語 or others and get the same plan and the same guardrails.

Token-credit billing per organization, with free monthly credits and every run showing its exact cost. No credit card required to start.

Put a human-gated agent in your ad ops.

Bring an account, write a goal in plain language, and watch the orchestrator plan the work step by step — then approve what ships.