Agent Personalization: an agent that knows you

Agent Personalization tailors every response the agent gives you. Instead of treating every user the same, Falconer builds a lightweight profile of who you are — captured as a set of attributes like your role, your team, the systems you own, the projects you care about — and uses them to tailor answers and skip context you already know. Your attributes are yours: you can see every one, edit it, or remove it.

What is an attribute?

Attributes come from two sources:

Attributes you add: facts you write directly in your profile — your role, your team, and personal preferences like “I prefer bullet points over tables.”
Auto-detected attributes: facts Falconer extracts from your Slack and in-app conversations as you use the product. Each one is traced back to its source so you can verify it.

Every attribute is reviewable from a single page. You can filter by “added by you” vs. “auto-detected,” search across attributes, edit any one inline, or delete what you don’t want Falconer to use.

How it works

As you interact with Falconer, the system:

Observes your conversations across Slack and the Falconer app
Extracts durable, identity-level attributes — not one-off questions or transient context
Traces each one back to its source so you can verify it
Injects relevant attributes into the agent’s system prompt at query time
Lets you correct or remove any attribute that’s wrong or outdated

Control and customization

Your profile is editable end-to-end. Add a role and team in one click. Add a freeform attribute up to 500 characters. Edit any auto-detected attribute — once you do, it’s marked “Edited by you” and Falconer trusts your version. Delete anything you’d rather Falconer forget.

Why this matters

Agent answers tuned to your role and ownership area, not a generic default
Less repetitive setup — you don’t re-explain your team or scope on every query
Full transparency: every attribute is visible, attributed, and reversible
Profile improves passively as you use the product, with no extra effort

Best use cases

Engineers and PMs who want answers scoped to the services they own, not the whole codebase
Non-technical stakeholders who need plain-language summaries, not implementation details
Writers and PMs who prefer a specific style — no em dashes, shorter sentences, bullet points over prose
Cross-functional teams where the same question means something different depending on the role asking it
Power users who want the agent to skip onboarding-level context and get straight to the point

Try it today

Agent Personalization is available now under your personal settings. Visit Settings → Personalizations → Agent Personalization to review your attributes.