Stop rewriting AI drafts. Start approving them.

Product pages in bulk, newsletters, campaign content — drafted by a team of specialists working from your rules, your catalog and your channels. Nothing goes out until someone on your team signs it.

The workspace mid-run: Oscar the orchestrator has split a launch brief across four agents, Kai has finished step four of four, and the completed LinkedIn post stands open in the canvas on the right

One place, not eleven tabs.

The brief, the agents doing the work, the draft, the image and the approval all happen in the same window. The conversation runs on the left; the thing being made stands open on the right. You watch it happen and step in wherever you want to.

Book a call See the pricing

Switch models, keep your work

Change the model mid-task and the thread, the context and the rules stay put. Pick a different engine halfway through a draft and nobody has to re-explain the brief — the work carries on from exactly where it was.

Visible reasoning

See which context, rule and agent shaped the answer. When a draft surprises you, you can tell at a glance whether the brief was off or the rule was, instead of guessing at a black box and rewriting on instinct.

Remembers what you tell it

Save a preference once and Tisser applies it everywhere from then on. The correction you make today stops coming back next week — it becomes a rule the next brief already knows about.

Edit with words

“Put her on a beach” and the whole scene follows — light, background and shadow move with it. No masks, no layers, no round trip for a change you can describe in one sentence.

Auto (system)
Model A
Model B
Model C
Thinking
RuleBrand contract · 4.2
AudienceMindful Millennials
AgentLynn
ActionDraft post
New preference
no dashes in bullet points
A studio portrait, before the edit The same portrait, moved to a beach at sunset

Inside the workspace

Everything the work needs, behind one door.

Four screens your team lives in. Same rules, same memory, same audit trail behind all of them.

The workspace context screen: nine cards covering brand, audience, product, rules, channels, proof, campaigns, competitors and sales play, each with a completeness figure

Your brand, written down once

Tone, audience, products, rules, channels, proof — the cards every agent reads before it writes a word. Each one shows how complete it is, so you can see what the work is standing on. One source, used everywhere. How the brand context works →

The model picker open in a new chat, listing GPT, Claude, Gemini and Nano Banana side by side with Auto at the top

Change engines, not rules

GPT, Claude, Gemini and Nano Banana in the same list — or leave it on Auto and let the workspace pick per task. Your rules sit above the model, so a model change never means a rules change.

The image editor: a prompt on the left, model and aspect ratio options beneath it, and the generated portrait on the right

Images in your visual language

Prompt, reference, aspect ratio, result — every format a channel asks for, out of one brief. The same brand rules apply to the picture as to the paragraph.

The projects overview, each project card showing its own chats and files and when it was last touched

A room per campaign

Every project keeps its own chats, files and context. The launch doesn’t leak into the newsletter, and picking a project back up in month four costs nobody an afternoon of re-explaining.

The Tisser mark at the centre of a ring of the tools a marketing team already works in, among them Slack, HubSpot, Salesforce, Notion, Figma, Google Ads, Mailchimp, LinkedIn and Outlook

Connect what matters to your business.

Your brand facts are not in one place. They sit in the CRM, the analytics, the drive, the calendar and the channels you publish to — and they change without anyone telling the AI.

Connect those sources to the workspace and every agent drafts from what is true today rather than from what someone pasted into a prompt last quarter. The finished work goes back out through the same tools your team already opens every morning, so nothing has to be copied across by hand and nothing lives in a chat window only one person can find.

What you get

Ten specialists, one set of rules.

Not one generic assistant. A team where every role has its own job description and its own memory — and the same brand contract binds all ten. You bring in the ones the work needs; Oscar decides which those are.

Portrait of Oscar, the coordinator agent

Oscar

Coordinator

Reads the request, decides which agents it needs and coordinates the run.

Portrait of Nora, the strategist agent

Nora

Strategist

Turns an objective into audience, positioning and measurable KPIs.

Portrait of May, the brief builder agent

May

Brief builder

Structures the chaos into one brief, one scope, one set of approvals.

Portrait of Lynn, the copywriter agent

Lynn

Copywriter

Performance-ready copy in your voice, checked against your claims.

Portrait of Max, the creative director agent

Max

Creative director

Translates strategy into a distinctive concept and steers it to approval.

Portrait of Rick, the designer agent

Rick

Designer

Scalable campaign visual systems and spec-correct assets, every channel.

Portrait of Kai, the social media manager agent

Kai

Social media manager

Platform-native social content and the community work around it.

Portrait of Vera, the market analyst agent

Vera

Market analyst

Detects market signals and turns them into evidence-backed intelligence.

Portrait of Rafa, the analytics lead agent

Rafa

Analytics lead

Turns raw signals into KPI frameworks and optimisation decisions.

Portrait of Tom, the account director agent

Tom

Account director

Owns scope, alignment and the cadence of client communication.

Meet the full team →

For IT and compliance

Your legal team will read this page too.

Data residency
All processing and storage in EU datacenters. Nothing leaves the EU. One contract, not forty personal accounts on personal cards.
GDPR
Standard DPA available before the first workshop. Data deleted on contract end, on a schedule you set.
EU AI Act
Article 50’s transparency obligations apply from 2 August 2026. Tisser One labels machine-drafted content by default — and because a named person reviews and holds editorial responsibility for every published piece, the human-review path through Article 50 is built into how the loop already runs.
Access
Named accounts per user. Approval rights assigned per person, per channel — so “who can publish what” is a setting, not a policy PDF.
Audit
Full activity trail, exportable, readable without Tisser present.

◻ AI-drafted · reviewed by Mara de Vries · 24 Jul 2026 — the label Article 50 asks for. The full security page.

Who runs it

In production, not in pilot.

bol. · Equans · Tony’s Chocolonely · Basic-Fit

[ customer quote — one, two sentences, answering the problem, not the product ]

Getting live

Two weeks from this page to a working team.

Software alone doesn’t change how a team works. We run the kickoff, connect your sources with you, train your approvers, and stay through the first live campaigns. That’s included. An honest word: in week one the drafts will need real editing — that’s not a flaw in the loop, it’s the loop. Your corrections are what make month three good.

Day 1

Sources connected

Brand memory intake — tone, catalog, channel rules.

Week 1

First draft in your queue

Agents draft against real briefs; approvers review.

Week 2

Team live

Full team running, audit trail from day one.

Setting up the brand contract is a named programme with a fixed scope and an end date. See the programmes →

Questions your team will ask

Before you book the call.

How do I keep AI content sounding like us?

You write down what that means once. Every draft is written against those rules, and a person on your team approves it before it goes out.

Who checks it before it’s published?

Someone on your team, by name. Anything going outside the company waits for them.

We already have brand guidelines. Do we start over?

No. Most brands start by writing down what already exists.

Which AI models does this use?

Whichever fits the task. Your rules sit above the model, so a model change never means a rules change.

Where does our data live?

In the EU, on ISO 27001-certified Google Cloud infrastructure. The certificate is Google's, for the data centres; we don't claim one of our own.

What does it cost?

One price per brand, per month — no calculator. What shapes it is on the pricing page.

How is this different from ChatGPT with a good prompt?

A prompt is a suggestion and it lives with one person. Rules are written down, apply to everyone, and the work stops for approval before it goes out.

Bring your own work. We’ll run it through the rules.

30 minutes, with a founder — not an SDR. Pick a slot in the calendar. Your data stays in the EU.

Want the same rules without us in the room? That’s Tisser Studio — closed beta.

Book a call