MaveraMavera Docs
Operator's Guide
Checking…
Use or share this pageSend the page itself, print it, or open its complete context in your AI workspace.

Workflows & Programs

Guided recipes, the natural-language builder, and long-running research.

2 min readDocs build Updated Jul 26, 2026
Interactive product walkthrough

Build a workflow

Describe a repeatable job, inspect generated steps and costs, then save it.

Simulation onlyNothing is saved or charged
1
Step 1 of 6Open Workflows

Choose Workflows from the sidebar.

Demo workspace / Mavera workspace
Local simulation
Mavera workspace

Good afternoon

Choose a surface from the sidebar to begin.

Demo workspace0 actions completed
WorkspaceDemo companySafe local simulation
Current surfaceworkflowFollow the highlighted action
Recent workspace activityStatus
Audience comparisonReady to review
Customer interview evidenceGrounded
Campaign pre-flightDraft
  1. 1
    Open WorkflowsChoose Workflows from the sidebar.
  2. 2
    Start a workflowCreate a workflow from a plain-language brief.
  3. 3
    Describe the recipeExplain the repeatable job and its required output.
  4. 4
    Build the planTurn the brief into ordered steps and gates.
  5. 5
    Review cost and stepsInspect inputs, estimated credits, and approval points.
  6. 6
    Save the workflowSave the local recipe so it can be reused.

Workflows: recipes with the price on the menu

The Workflows page has ready-made research chains — positioning test, pricing read, persona deep-dive, and more. Each card shows estimated credits and minutes before you run it. Fill 2–3 fields and go; the run lands in the same Mave conversation view you already know.

The builder: describe it, get a plan

At the top of Workflows: Build your own workflow. Type your goal in plain language:

"Figure out if suburban parents would switch to an electric minivan under $50k, and which feature to lead marketing with."

Mave designs the step chain — research the landscape, build the audience, field the survey, run the focus group — with a cost and time estimate. You review the steps, then one click runs the whole thing. Nothing is charged until you approve.

Programs: research that keeps working

Programs are for the big asks: a market map, a GTM blueprint, a competitive landscape. A program plans a step, executes it, reads what it found, and re-plans — for up to dozens of iterations, pulling in cited web sources as it goes. Come back in an hour to a structured report with sections, charts, and links to every source.

Programs can be paused, stepped manually, and steered from their detail page. The Logo Lab tab (on GTM programs) generates brand marks in batches — ❤️ the ones you like and the next batch leans toward them.

SOW executions: contracts become evidence gates

For a structured naming contract, use an SOW execution. The compiler turns the source text into a 26-item requirement ledger and a 15-module dependency graph. You bind the exact candidate names and a fieldable audience, approve the contract, then start the worker-backed fieldwork.

SOW executions differ from ordinary programs in one important way: completion is contractual. Every required module needs its respondent count and evidence references. Zero-row jobs, missing lineage, underfilled modules, and missing dependencies stop the graph instead of producing a polished but unsupported result.

See SOW executions for the API, fieldwork plan, DAG, and runnable request consoles.

Which one when?

  • Cost certainty and speed → Workflow
  • "Go do the research and come back with a document" → Program
  • Just want an answer → ask Mave, she'll route it

Running studies

Structured surveys with numbers you can defend.

Context, brand & memory

Teach Mavera your business once — every answer gets smarter.

On this page

Workflows: recipes with the price on the menuThe builder: describe it, get a planPrograms: research that keeps workingSOW executions: contracts become evidence gatesWhich one when?