osFoundry / Platform
Records
A structured table view for any data model — customers, invoices, leads. Like a database with eyes.
What it is
Records is osFoundry's structured table view: a place to model any set of business entities — customers, invoices, leads, employees, vendors, products, assets, expenditure, income — as typed rows and columns rather than scattered notes or spreadsheets. You start from one of 18 built-in templates that seed a sensible schema (a Customer table arrives with segment, lifecycle stage, lifetime value, owner and health-rating columns already defined; an Income table with source type, amount, tax, payment method and status), or you open a blank Custom grid and define the fields yourself.\n\nUnlike a plain spreadsheet, every column has a real type. Text, number, date, boolean, email, phone and URL are the basics; on top of that you get currency, percentage, single- and multi-select dropdowns, star ratings, file attachments, relations that link one table to another, and formula columns. A field can be marked required so rows aren't saved half-empty. You can import a CSV or spreadsheet and map its source columns onto your fields, so existing data lands correctly instead of as an untyped blob, and you can reshape the schema afterward from the Schema tab. Tables can hold data manually, reference a live source, or snapshot it.\n\nWhat makes it 'a database with eyes' is that Records sits inside the same workspace as your chat and agents. A table can be attached to a conversation as context, so an AI model reasons over your actual rows — summarize this quarter's income table, flag leads that stalled, cross-reference vendors against expenditure — using whatever model you've brought your own key for. Because osFoundry is a local-first desktop app, sensitive tables like customer lists or bank-account records can stay on your device, and there are no per-seat fees for having a database.
Key capabilities
- Build a structured table for any data model — 18 starter templates (customers, invoices/income, leads, employees, vendors, assets) or a blank Custom grid you shape yourself
- 16 real column types: text, number, currency, percentage, date, email, phone, URL, single/multi-select, rating, file attachments, relations between tables, formulas, boolean, and JSON
- Import a CSV or spreadsheet and map its columns onto your schema, so existing data lands in the right fields instead of a flat dump
- Records feed your AI: any table can be attached to a chat or agent as context, so a model reasons over your actual rows — the 'database with eyes'
- Runs local-first in the desktop workspace with your own keys — your customer and financial tables can stay on your device, no per-seat database fees
Use cases
- Ops lead: build a Customer table from the template, import an existing CSV of accounts, then ask an AI to flag which accounts are near renewal
- Founder/finance: track Income and Expenditure as typed tables with currency, tax and status columns, then have a model summarize the quarter
- Recruiter: run a Candidate pipeline table with stage, source, rating and resume attachments, and let an agent shortlist by criteria across the rows
Frequently asked questions
How is Records different from a spreadsheet like Excel or Google Sheets?
Every column has an enforced type — currency, date, select dropdown, rating, file, relation, formula — instead of loose cells, so a currency column stays a currency and a select column only accepts its defined options. You also start from templates with a ready-made schema, can require fields, link tables together by relation, and attach a whole table to an AI model as context. It's closer to an Airtable-style database than a freeform sheet.
Can an AI actually read and reason over my Records data?
Yes. That's the point of the feature living in the workspace. You can attach a table to a chat or agent, and the model reasons over the real rows — summarizing, filtering, cross-referencing — using your own API key (or a local/self-hosted model). The AI works from your structured data rather than you copy-pasting it into a prompt.
Where is my Records data stored — is it in the cloud?
osFoundry is a local-first desktop workspace, so your tables can stay on your device rather than a vendor's server. That matters for sensitive tables like customers, bank accounts or income. There's no per-seat database charge, and because the whole product is bring-your-own-key, no data is routed through a reseller to use an AI on it.