Build your own AI team. You stay in charge.
Start with a template, such as a research analyst, a strategist or a reporting analyst, or describe in plain words what you need. Your team works on your brand DNA and puts everything to you for approval.
Templates or your own
Start with a template, or build your own agent.
Pick a template, or describe what your agent should do and build it together with the assistant. You choose a starting point, a persona and a schedule, and you check everything before it goes to work.
Every change the assistant proposes, you confirm yourself.
Skills and context
Skills that have been measured, and your documents as fixed knowledge.
A skill only enters the Branddock library after it has been measured. Per agent you choose up to ten of your own documents that it always brings along.
Anything that doesn't make a measurable difference doesn't make it in.
Rooms and team discussion
Work with your agents the way you work with colleagues.
Open a room per campaign or topic. Let your agents talk it through: after the discussion you get a report and their proposals.
Everything that needs your decision is under Waiting on you.
Rhythm
On a set rhythm, or when something happens.
Daily, weekly or monthly. Or when something happens: when a review is requested, when content is published, or when a connected tool finishes syncing.
In developmentIn a loop, Branddock plans the work itself, at the pace of your measurement source.
Events that fall within five minutes, Branddock bundles into one task.
Control
Budget, access and a stop button.
Each agent has a monthly budget in credits. Per connector you decide whether the agent only reads or also acts. You stop a running task with one button, after the step it is working on. And you can approve from your email.
The approval link in your email works once and stays valid for 24 hours.
Strategy
Campaigns with a foundation.
The strategist builds a campaign direction on behavioral science and effectiveness research, not on a one-off prompt. You then work out the full campaign in the campaign wizard.
Drawing on the work of Binet and Field, the principles of Ehrenberg-Bass and the EAST approach, among others.
What it doesn't do (yet)
Agents publish nothing on their own: whatever they make or want to change is a proposal for you. When an agent's monthly budget runs out, it starts no new task. They do not run continuously, but only when you start them, at the pace you choose, or on an event. An agent on a schedule does not learn from what it did. That is what loops do, and those are in development.
Frequently asked questions
How many agents can I create?
As many as you like: there is no maximum on the number of agents. What they do runs through your credits and the monthly budget per agent.
What does an agent cost?
A task costs credits based on the length of the result, with a guide value of 3 credits. A task that starts on an event costs a fixed 3 credits.
Can an agent search the internet?
Yes, if you turn on web search for an agent. It then searches at most 3 times per task, and each workspace has a default ceiling of 25 searches per day. That search function costs no credits. In addition, some templates, such as the research analyst and the strategist, do their own source research as part of the task.
Discover the rest of Branddock
- Brand DNAYour brand, captured once.
- Brand checkA brand score with an explanation for each finding, on campaigns, pages from the app and anything you submit for a check.
- ConnectorsWorks in the tools you already use.
- Publish and measureOut into the world, and see what it does.
- LoopsIn developmentWork toward a goal, and learn what works.
See it on your own brand.
Start with your own website: Branddock turns it into the first version of your brand DNA. Then you sharpen what only you know.