Since the release of the Studio agent, adding context has been key to getting output that sounds and feels like your brand.
It's here now!
🗂️ Overview of context layers
Studio uses five layers of context to inform its suggestions. They are described here:
# | Layer | How it's applied |
1 | Platform level | Auto-injected |
2 | Webshop | Auto-injected |
3 | Activity (via document upload) | Auto-injected |
4 | Neighbouring activities within a programme/journey | On demand |
5 | Assets | On demand |
In definition
Auto-injected: Always active, no action needed
On demand: Only used when you ask for it in your prompt.
⚙️ Layer 1: Platform-level context
Platform-level context applies across your entire platform and is automatically passed to the agent on every request.
For example:
The best place to add your company mission, writing tone, and brand-specific terms!
To set this up:
Go to Settings > General.
Scroll to Context for the Studio agent.
Enter your platform-level context and save.
Don't see this option?
Contact robin.seegers@tinqwise.com: a signed AI addendum is required before it can be enabled.
No data is used to train the AI and your information is kept secure.
🛍️ Layer 2: Webshop
If you have a connected webshop, Studio automatically pulls in up-to-date product information so the agent always works with accurate data.
Tip:
A connected webshop also unlocks the Product trainer, an AI-generated quiz based on your products. Contact robin.seegers@tinqwise.com to get set up!
📄 Layer 3: Document upload within an activity
You can upload a document directly inside the Studio agent while working on an activity. The agent will read the document and use it to suggest modifications and additions to your content.
For example:
This is useful for any reference material relevant to that specific activity.
🔗 Layer 4: Neighbouring activities in a programme
When working on an activity that sits within a programme, you can ask the agent to reference other activities in the same programme.
For example:
"Use the other activities in this programme to keep the tone consistent."
🖼️ Layer 5: Assets
The agent can also reference documents stored in your assets library. Like neighbouring activities, this is fetched on demand, you need to ask for it in your prompt.
For example:
"Look for the onboarding script in assets and use it as a reference."