Account & preferences

Your account page is where you set your display name and email, and tune how the platform talks to you specifically. The "Reasoning preferences" panel is the most useful part — it changes how every answer in the current project is composed.

Where to find this

Account menu (avatar top-left) → Account, or go to /settings/account directly.

What's on the Account screen

Profile section

• DISPLAY NAME input (defaults to your name from Google).
• EMAIL input.
• Avatar (auto-derived from your initials or Google avatar).

Your name and email are editable; the underlying authentication is Google OAuth, so changing your email in Google updates the identity. Changing it here only changes how the platform displays you.

Reasoning preferences section

"How you want the agent to talk to you. Applied to every reasoning answer in the current project; missing fields fall back to defaults. Changes take effect on the next question."

These are scoped to the current project (each project can have a different tone for the same user). Available controls:

• YOUR ROLE — let the platform know who you are (Growth Marketer, Finance Lead, Founder, etc.). Defaults to "No preference". When set, the agent biases its framing — e.g. CFO context produces tighter unit-economics language; Growth context surfaces channel-level detail.
• VERBOSITY — Terse / Standard / Detailed. Default "No preference".
• DEFAULT TIME TENSE FOR METRICS — "as of latest", "trailing 7 days", "month-to-date", "quarter-to-date" etc. Default "No preference". When set, the agent picks this window by default when you don't specify one.
• CURRENCY DISPLAY — INR / USD / Auto (mixed). Default "No preference".
• GLOSSARY OVERRIDES — + Add rows that override how a term gets resolved in your project (e.g. "NRoAS" → "New-customer revenue / total spend"). These are personal overrides — they don't affect other people's runs on the same project.

Memory / "forget" section

Lower on the page, a list of facts the platform has learned about your preferences from past interactions ("you prefer monthly cohorts over weekly", "you typically work in INR"). Each fact has a Forget button to remove it.

What's on the Profile screen — "About you"

/settings/profile is a separate page from Account. While Account is project-scoped reasoning preferences, Profile carries persistent user context that applies across every project — your role, the function you work in, your team, your scope, the topics you keep coming back to.

Header: "About you" — "The system reads these fields before answering your questions. Setting your role + function lets it tailor depth and emphasis."

Fields:

• ROLE — free-text job title (e.g. "Founder / Builder", "VP Marketing", "Data Scientist").
• FUNCTION — dropdown. Available functions match the project functions (sales, marketing, growth, customer, product, people_org, decision_intelligence, leadership, etc.).
• TEAM — free-text team name (e.g. "Founding", "Growth", "Brand", "Demand Gen").
• RESPONSIBILITIES — textarea describing what you own. The agent uses this to frame answers in terms of your scope (e.g. a CFO context gets unit-economics framing, a Growth lead gets channel-level detail).
• PREFERRED RESPONSE DEPTH — dropdown: terse / balanced / detailed. Tip from the UI: "Terse for power users, detailed for stakeholders, balanced for everyone else."
• DOMAIN FOCUS — chip-based tag list of topics you frequently return to (e.g. decision_intelligence, extensibility, creative_concepts, cac, attribution). Each chip has an ✗ to remove. Below the chips: a free-text input + Add button to drop in new topics. The platform uses these tags to bias retrieval of past corrections and memory facts when you ask a question.

Account vs Profile — which is which

Set Profile once when you first sign in. Tune Account per project as your work shifts.

Tips

• Set YOUR ROLE per project if you have hats. CFO context for the Finance project, Founder context for the strategic project, Operator context for the daily-driver project.
• Set VERBOSITY to Terse on your morning project, Detailed on the deep-dive project.
• Avoid over-overriding glossary terms. Personal overrides drift from canonical; over time they make your answers inconsistent with your team's. If a term is wrong project-wide, ask the DA to fix the ontology, not your overrides.
• Periodically Forget stale facts. What was true six months ago may not be true now.

Related

• Workspaces & members — workspace-level role
• Permissions
• Ontology — where the underlying glossary lives