Platform
Solutions
Customers
Resources
Pricing Book a demo
Platform · Studio

Studio

The visual canvas where marketing teams build agentic workflows. Drag nodes onto the canvas, connect outputs to inputs, and pass data between steps with variables — no code required.

The Studio builder — drag nodes from the library, connect output → input, run.
Capabilities

Everything you need on one canvas.

Drag-and-drop canvas

Drag nodes from the library onto the canvas and draw connections from one node's output to the next node's input. Pan, zoom, multi-select, and duplicate — your whole workflow laid out visually.

A node for every step

Input, AI, Integration, Tool, and Control-flow nodes — from short-text and file inputs to LLMs, web scrapers, conditional IF/ELSE branches, and merge-to-report.

AI agents & LLMs, built in

Drop in an AI Agent that plans, calls tools, and adapts — or an LLM node with structured JSON output. Choose GPT, Claude, or Gemini per node, and attach RAG datasources and MCP servers.

Native SEO & research tools

First-class nodes for Semrush, DataForSEO, YourText.Guru, Haloscan, Google SERP, Search Console, and BigQuery — plus a Web Scraper and API Connector to reach anything else.

Shape data without code

Transform between steps with the JSON Path Extractor, HTML-to-Markdown, Tag Extractor, Merge, and Find & Replace — and an AI Code node for the moments where no-code shouldn't go.

Variables connect everything

Pass any node's output into the next with {{variables}}. Reference an upstream result anywhere in a prompt or parameter to wire data cleanly through the workflow.

The canvas

Your whole workflow, laid out visually.

The canvas is the main workspace. Drag nodes from the library, draw connections from a node's output to the next node's input, then pan, zoom, multi-select, duplicate, and delete to shape the flow — selecting any node to configure it in place.

Full-width product screenshot of the workflow editor.
Node library

Four categories. A node for every step.

Input & output01

Capture what a run needs and hand back a finished asset.

Short TextParagraphURLDateFileDownload File
AI02

Reasoning, generation, and analysis — pick a model per node.

AI AgentLLMAI CodeImage to TextText to ImageData Analyzer
Integrations03

Pull live data in and push results out.

Web ScraperAPI ConnectorGoogle SheetsBigQuerySearch Console
Tools04

Reshape data between steps — no code required.

JSON Path ExtractorHTML to MarkdownTag ExtractorMergeFind & ReplaceInternal Link Recommendation
AI agents & LLMs

A reasoning runtime, not just a prompt box.

AI Agent node

Add an AI Agent that plans, calls tools, retrieves context, and adapts across multiple steps. Pick a reasoning-capable model — GPT-4, Claude Sonnet/Opus, or Gemini Pro — then attach Tools, RAG datasources, and MCP servers, and tune temperature, top-K, top-P, and max tokens.

Model per node

Choose the right model for each task — no platform lock-in.

Tools & MCP

Let the agent drive web search, scrapers, and your own MCP servers.

RAG datasources

Ground answers in private knowledge bases you attach.

Structured output

The LLM node returns JSON to a schema for clean downstream use.

Agent runtime
The agent loop
A reasoning runtime that plans, calls tools & datasources, observes, adapts — then answers.
iterate until confident
REASONING RUNTIME
01
Plan
02
Call tool
03
Observe
04
Adapt
05
Answer
invokes
Tools
External
web.searchcode.runhttp.fetch
Datasources
External
vector.dbsqldocs
Native SEO & research tools

The marketing stack, as first-class nodes.

No HTTP plumbing for the tools marketers live in — they're nodes you drag onto the canvas. And for anything without a native node, the Web Scraper and API Connector reach the rest.

Semrush Domain Keywords
Semrush Keyword Overview
Semrush URL Keywords
DataForSEO
YourText.Guru
Haloscan
Google SERP
Search Console
Google Discover
Google Ads
Web Scraper
API Connector
Variables

Wire data through with {{variables}}.

Every node's output is reusable. Reference an upstream result anywhere in a prompt or parameter with a variable, and Studio threads the data through the workflow — no glue code, no copy-paste.

Example

An SEO content workflow, end to end.

Keyword in → reviewed draft out
Input: keyword Semrush Keyword Overview AI Agent: build brief LLM: draft (JSON) Human-in-the-Loop Google Sheets

Each step passes its output to the next as a variable. The Human-in-the-Loop node pauses for an editor to approve before anything is written back.

How it works

From blank canvas to production.

01

Build

Start a new workflow, drag in an input or data-source node, then add AI, tool, and integration nodes — connecting outputs to inputs as you go.

02

Test

Run the workflow right in Studio and inspect every step's output on the canvas before you ship it.

03

Run at scale

Trigger runs from the API or on a schedule, and batch over a CSV or Google Sheet. Runs execute asynchronously, with per-step status you can monitor.

04

Review & resume

Add a Human-in-the-Loop node to pause for approval. A reviewer validates the content, and the workflow resumes from exactly where it left off.

A run in progress — completed, running, and failed steps.
Get started

Show us the workflow.
We'll show you the 10x.

Bring the marketing workflow that eats your week. We'll build it live, with your data and your models, in 30 minutes.