Data Handling

Kimpton AI handles three main categories of your data: portfolio holdings synced from your brokerage, documents you upload to the Vault, and the research sessions and outputs you generate on the platform. Across all three, the core commitment is the same: your data is encrypted, isolated to your workspace, and never used to train Kimpton's AI models. The sections below explain exactly how each category is stored and protected.

---

Portfolio data

How Kimpton connects to your portfolio depends on your plan class. Kimpton has three: Free, paid and Enterprise. Every paid plan behaves identically; they differ only in how many credits they include, so there is no capability you get on one paid plan and not another.

---

Vault documents

The Vault is where you store private research documents, such as mandate files, internal memos and third-party reports, that you want available as AI context.

• Encrypted at rest with AES-256 on Kimpton's infrastructure.
• Private by default, shared deliberately — a Vault item belongs to you and nobody else can see it until you share it. You can then share an item with specific people, with a team, or with your whole organization. Nothing is shared implicitly, and sharing is per item rather than all-or-nothing.
• Used only when you enable Vault context — Kimpton does not silently inject Vault documents into every session. Documents are retrieved as context only when you explicitly enable Vault for that session.
• Never used for model training — your uploaded documents are not used to fine-tune, retrain, or improve any AI model.
• Deletable at any time — you can remove any document from the Vault at any time through the Vault interface. Deletion is immediate and permanent.

---

Research and conversation data

Research sessions and their outputs are stored to support your research history and session continuity, so you can return to a prior analysis or pick up where you left off.

• Research sessions are scoped to your account. If you are working within a team workspace, sessions created there are scoped to that workspace and are not visible to users outside it.
• Session data is not used for model training under any circumstances.
• Research outputs (summaries, memos, analyses) you export or save are treated with the same isolation guarantees as other workspace data.

---

Third-party data providers

Kimpton routes market data queries through licensed API integrations with institutional market-data providers. A few important points:

• Your queries are transmitted to these providers to retrieve relevant data. Each provider's own data usage policies govern how they handle query data on their end.
• Where you connect your firm's own licensed data subscription, Kimpton authenticates via OAuth and uses a scoped OAuth token. It does not receive or store your data-provider account credentials.
• Kimpton does not sell or share your query content with data providers beyond what is necessary to fulfill the data request.

---

Data retention and deletion

For full retention schedules, lawful bases for processing, and your rights as a data subject, refer to:

• Privacy Policy
• Data Processing Agreement