The same kind of intelligence,
a different way of working.
A general-purpose AI chat is an extraordinary tool: it answers, writes, reasons, and today it can also search the web and work on the files you upload. MAPS uses models from the same family — Anthropic’s Claude, and GPT-OSS-120B, OpenAI’s open model.
The difference is what surrounds the model: a project memory that grows and stays separate for each client, sources and dates on every piece of information, strategic methods already built, synthetic users to put ideas to the test, and the Studios to produce the work.
What changes,
in everyday work.
Memory
The conversation, the files you upload, what the chat remembers about you.
The project’s memory: research, analyses, material, decisions. It updates with every save and every module reasons on it.
Sources
It depends on the question and on what it finds at that moment.
Every piece of information carries the module it comes from and the date; news reads cite the source and date of every statement.
Research language
The one you write in, unless you ask otherwise.
The language of the market you’re studying: a market is read in its own words.
Methods
You bring them, with the right prompt.
Already built: SWOT, positioning, value proposition, customer journey, semiotics, canvas lenses and flows, creative methods.
Putting it to the test
You can ask for an opinion.
Synthetic users built on the project’s data and the Buying Committee: declared simulations, so you arrive prepared.
Multiple clients
Keeping contexts and files separate is up to you.
Each client has its own space; when you reason across several clients or countries at once, every piece of information carries its label.
From thinking to work
Text to copy wherever it’s needed.
The Studios produce posts, emails, web pages, videos and presentations consistent with the brand, and they go through client approval.
For a quick question,
it’s perfect.
A one-off question, a draft, a brainstorm with no project behind it: a general-purpose AI chat is the right choice, and it costs less.
MAPS is for when the work is ongoing: a market to follow over time, several clients, decisions to justify, content to get approved.
MAPS inside
your AI chat.
With MCP, Claude and ChatGPT connect to MAPS: they read the research and analyses in your projects and launch new analyses. They answer with MAPS data, not just the web.
The questions
people ask us.
Is MAPS an AI chat?
No. MAPS has a chat, but it’s a marketing intelligence platform: it reads the market, helps you decide the strategy, puts it to the test and turns it into work. The chat is one of the ways to query what it knows.
Which models does MAPS use?
Anthropic’s Claude for strategic reasoning and production, GPT-OSS-120B (OpenAI’s open model) on Groq to structure data and research, and OpenAI for semantic search in the memory.
Can I use MAPS from Claude or ChatGPT?
Yes, with MCP: you connect MAPS as a connector and your AI chat reads your projects and launches analyses.
Is my data used to train the models?
No. The model providers receive the text via API and, by contract, don’t use it for training.