Master four concepts — Projects, Models, Credits, and Context — and every Studio session becomes faster, cheaper, and more predictable.
The AI automatically selects tools and manages context within a session. Understanding these concepts lets you guide it more effectively, not replace its decision-making.
Equip users with the mental model needed to structure their work: scope a project, choose the right model tier, manage credit spend, and keep long conversations productive.
- Signed in to Treasure AI Studio (Getting Started)
- AI Credits provisioned for your account
A project saves your working context so you don't repeat yourself across conversations. Think of it as a preset that loads your preferences and skills every time you start a new chat.
Each project includes:
| Field | What It Does |
|---|---|
| Name | Identifies the project in the sidebar and project switcher |
| Directory | A slug (lowercase letters, numbers, hyphens) used internally. Cannot be changed after creation |
| Instructions | Custom rules the AI follows — coding conventions, preferred databases, output formats |
| Skills | Skills that are always active in this project's chats |
| TD context | Default database, table, or segment the AI uses when none is specified |
Example: A "CDP Ops" project might activate segment and journey skills and include instructions like "Always validate YAML before pushing. Use the marketing database by default."
Projects appear in the PROJECTS section of the sidebar with a count in parentheses (e.g., "PROJECTS (3)"). Click the folder-plus icon next to the header to create a new project — a Create Project dialog appears asking for Name and Directory.
Projects appear on the home screen in a collapsible PROJECTS section. Each project row shows its chat count. Tap a project to view its chats, or tap the floating action button to start a new chat.
Project configurations are stored server-side and sync across all your platforms. A project created on Web is immediately available on Desktop and Mobile.
Studio supports three AI model tiers. You can switch models mid-conversation — useful for starting with Fast for quick exploration, then switching to Pro for the complex part. Switching models restarts the AI session. Your conversation history is preserved, but the AI re-initializes with the new model.
| UI Label | Model | Description (from UI) | When to Use |
|---|---|---|---|
| Pro | Claude Opus 4.6 | "Complex reasoning and analysis" | Multi-step data analysis, building segment rules, writing workflows |
| Balanced | Claude Sonnet 4.6 | "Fast and capable" | General-purpose work — the default model for new conversations |
| Fast | Claude Haiku 4.5 | "Quick tasks and simple questions" | Quick lookups, simple questions, conserving credits |
Click the model selector dropdown in the chat input area to switch. The current model label (e.g., "Balanced") is shown.
Tap the model label in the chat input area to switch between available models.
Switch to Fast for simple lookups like "list my databases" or "How many rows in this table?" — it answers in seconds and uses far fewer credits. Reserve Pro for multi-step analysis, complex segment logic, and workflow generation.
Each AI interaction consumes credits. The amount depends on which model you use and how much the AI processes (longer conversations and tool-heavy sessions cost more).
| Model Tier | Relative Cost |
|---|---|
| Pro | Highest |
| Balanced | Medium |
| Fast | Lowest |
Contact your Customer Success Manager to check your credit allocation or request additional credits.
The context window is how much conversation the AI can "remember" at once. As you chat, the context fills up. When it gets close to the limit, Studio's auto-compact feature kicks in — it summarizes older messages to free up space without losing important context. Your conversation continues normally.
You can also manage context manually with these commands in the chat:
| Command | Action |
|---|---|
/context | See detailed context usage breakdown |
/compact | Summarize the conversation to free up space |
/clear | Clear the context entirely and start fresh |
For best results on long-running work, use projects to carry context across multiple chats rather than pushing a single conversation to its limit.
Within a single conversation, the AI remembers everything discussed — questions asked, data explored, decisions made. This in-session memory lets you build on previous messages naturally:
- "Now filter that to just the US region" (references a previous query)
- "Add those users to the segment we defined earlier" (references a previous segment definition)
- "Export that as a PowerPoint" (references a previous chart or analysis)
In-session memory (the AI's context window) does not persist across separate conversations. However, the AI can read and write memory files that persist across sessions within the same project. Project instructions also carry forward automatically. See In-Session Memory for details.
All conversations are saved automatically and sorted by most recent activity.
Chats appear in the sidebar — under their parent project in the PROJECTS section, or in the standalone CHATS section. The search bar (placeholder: "Search chats...") filters chats by title. Use Cmd+O to open chat search.
The home screen shows a Recent Chats entry with a count of standalone conversations. Tap it to see the full list. Chats within a project are visible when you tap into that project.
Chat history is stored server-side, so your conversations are available on any platform you sign into. Treasure Studio Labs stored chats locally on disk — this is no longer the case.
Studio automatically generates a descriptive title for each conversation. Titles appear in the sidebar chat list to help you find past conversations.
To rename a conversation, right-click a chat in the sidebar and select Rename Chat. To delete, select Delete Chat.
Long-press a chat to access rename and delete options.
Access settings from the user avatar in the sidebar footer (Web/Desktop) or the user icon in the top-left header (Mobile).
Settings open as a centered modal with tabs on the left sidebar:
| Tab | Label | Contents |
|---|---|---|
| General | General | Language (English, 日本語), Timezone (searchable picker), Auto Update toggle (desktop only) |
| Appearance | Appearance | Theme preview cards: Light, Dark, System |
| Profile | Profile | User email, identity provider, region + Account ID, Sign Out button |
| Network Audit Log | Network Audit Log | Network activity log viewer with time range and filters (admin-only, under "Organization" section) |
The bottom-left corner shows the app version: "Treasure AI Studio v0.1.0".
Settings open as a bottom sheet (drag-to-dismiss). Tap a row to navigate to its sub-page:
| Row | Icon | Contents |
|---|---|---|
| Language | Globe | English or 日本語 (checkmark on current selection) |
| Appearance | Moon | System, Light, or Dark (checkmark on current selection) |
| Timezone | Clock | Searchable timezone picker |
| Sign Out | LogOut (red) | End your session |
The footer shows the app version.
- Create a project with instructions and an active skill
- Switch between Pro, Balanced, and Fast models during a conversation
- Use
/contextto check your context usage and/compactto free space
| Issue | Solution |
|---|---|
| Cannot create a new project | Verify you have an active Treasure AI account with AI Studio access. On Mobile, projects are managed on Web/Desktop |
| Model selector not visible | The model selector appears in the chat input area on Web and Desktop. On Mobile, the default model is used automatically |
/context returns no output | Context tracking requires an active conversation. Send at least one message before checking context usage |
- Getting Started — Set up Studio on your platform
- Supported Skills Catalog — Browse available AI skills
- Security — Configure permissions and security settings