Skip to content

Data handling

Last updated: August 14, 2026

What this page covers

This is a practical description of how research content moves through the product. It is not a SOC 2 report and not a claim of end-to-end encryption. See also security and privacy.

Where content lives

Uploaded files, parsed text, embeddings, chats, grids, and generated outputs are stored in our Supabase project (Postgres and object storage). The application runs on Vercel. Access is gated by authentication and row-level security.

What happens when you run research

The application retrieves relevant document excerpts and sends those excerpts, plus your prompt, to a model provider (OpenAI or Anthropic) to generate a response. Embeddings used for retrieval are created with OpenAI. We do not use your documents to train our own models. We do not currently operate a contractual zero-retention arrangement with model providers.

Public market sources

SEC filings imported from EDGAR are public. Earnings-call transcripts, when imported, come from Financial Modeling Prep (FMP), not from EDGAR, unless a transcript is itself an SEC-filed exhibit. Coverage news and related alerts may be fetched from third-party news APIs (including Finnhub, when configured) when a ticker is attached to a workspace. In-app prices, when shown, come from FMP and may be delayed. Marketing examples are labeled as illustrative when they are fictional.

Sharing

A workspace owner can invite editors and viewers. Those users can see the workspace content according to their role. There is no org-wide admin plane, SSO, or VPC isolation today.

Retention and deletion

You can delete documents and workspaces in the application. There is not yet a self-serve account deletion control. To request deletion of an account and associated research data, email contact@institutionalanalytics.com.

Billing

Pro subscriptions are billed per user through Stripe from account settings. The marketing site does not collect payment cards.