> SpecStream is the live web data engine for teams and agents. This is the Markdown copy of https://specstream.ai/docs/workspace. Every page, with what SpecStream does and costs: https://specstream.ai/llms.txt

# Visual workspace and saved work

## Choose a view.

Imported lists and saved results have Table, List, Map, Compare, Relationships, and History views. Search, filter, sort, and select records. Use table column controls to focus on the details you need. The Columns menu shows how many records in view have each detail, and **Fill the missing** adds that detail to the ones without it, after a quote. Select two or more records to compare their fields side by side.

Add your own fields to keep track of a list in SpecStream. Choose **+** at the end of the table header, or **Columns** › **Your fields**, and add a Status with options you name, a Checkbox, or a Text field. Fill them in from the table or the record panel, or select records and choose **Set** to fill many at once. Filter and search by them like any other column, and they go out with exports and to connected tools. SpecStream never fills in or changes your fields, and share links leave them out.

Save a view to keep its filters, columns, sorting, map area, and view type. Star a collection to keep it in the sidebar.

In Compare, turn on **Differences only** to focus on what changes between the selected records. Larger selections stay available across pages.

To make one list from several, select two or more collections on the Collections page and choose **Combine**. A place or company that’s in more than one of them keeps a single record, and the others fill in the details it doesn’t have. The original collections stay as they are.

Press ⌘K (Ctrl K on Windows) to search everything at once: records inside your collections, people, notes, monitors, snapshots, templates, and settings. It also starts things, like “Import from Google Sheets” or “Connect HubSpot”.

Open a record for its details, sources, and notes. Each value keeps its source URL and the time it was read. Your selection can be the scope for adding details or rechecking records.

## Explore locations.

The map uses each record’s coordinates: supplied with it, found for it, or placed from its address. Switch between clustered points and a heat layer, open a point, select records, or fit the map to the collection or selection.

Color the pins by match status, or by a column with a few values, such as category or neighborhood. The legend shows each color and how many records have it.

Filter the collection to the map area on screen. The map shows how many records are in view. Records without coordinates stay in the collection, listed separately.

Select two or more places and choose **Plan a route** for a visiting order through up to 11 of them, starting wherever you choose. Open it as walking, cycling, transit, or driving directions in Google Maps, or copy the stops.

## Follow the work.

The Relationships view connects records to values in your columns. Choose a field such as company, location, category, or supplier, then use the graph or table to explore shared values. This works with imported lists, too. Records with a missing value stay unconnected.

Under its name, a collection links the saved request that found it and every monitor that rechecks it. When a recheck finds something different, the same line says how many details changed. Open it to see each record’s value before and now, and where the new value was seen. **Activity** lists what happened across the workspace: runs with their outcome and charge, imports, snapshots, new monitors, and records sent to tools. Filter it to collections, requests, monitoring, tools, or projects.

The History canvas connects a starting request or import, the working collection, saved snapshots, follow-up research, and scheduled work. Move nodes, use the minimap, open related work, or switch to the timeline. The canvas layout is saved in your browser.

The timeline includes individual runs, their outcomes and charges, and scheduled work that couldn’t start. Open an entry to inspect the run or its monitor history.

Request revisions and the history between runs are saved to your account. Compare revisions and restore an earlier one into the editor as a new revision. The connections show work that happened.

## Import a starting list.

Import a CSV or Excel file with up to 20,000 rows and 20 MB, or paste a link to a Google Sheet, an Excel workbook in OneDrive or SharePoint, or an Airtable table or view. Linked lists can stay up to date: SpecStream checks the source every day for new and changed rows. Map the name, website, country, city, and coordinate columns, and retain other supplied columns. You can view, map, compare, export, and organize the list without starting paid work. Imported values stay labeled as yours.

Rows with an address and no coordinates are placed on the map for you, free, each at its postcode. To look up selected imported rows, each needs a website, a page link, or an ID such as a company number or DOI. Review the new request and quote before running it. You can add details or recheck the saved results after that lookup completes.

## Save and export.

Save a fixed snapshot of selected research results or imported records. Export CSV, Excel, JSON, or HTML. When the records have coordinates, export KML for Google My Maps and Google Earth, or GeoJSON for any mapping tool. Or send selected records straight to a [connected tool](https://specstream.ai/docs/integrations.md). The Excel file has a sheet each for Matches, Evidence, Details, Field guide, About this list, and Acquisition. Each download link works once. Reading or exporting work you own doesn’t buy it again.

## Share a selected snapshot.

In a collection, **Share** saves the records in view, or the ones you selected, as a snapshot, creates the link, and copies it. Choose an expiry from 1 to 365 days (90 unless you pick), whether recipients can download it, and whether contact fields are included. Every link stays listed with its records and expiry, so you can copy, open, or revoke it later. People who open a shared snapshot don’t get access to your account.

When SpecStream found the list, the link can also show the request that found it. People who open it see your exact words and **Run this yourself**, which opens the request in their own workspace, after they sign up if they’re new. Nothing runs until they approve a price.

Each link has **Copy embed code**. Paste the code into your own site to show the map with the list’s name and how many records it has, for as long as the link lasts. Visitors can open a place on the map to see its details.

You can also share a reviewed request revision as a reusable link, with optional context and an expiration. The recipient can save their own copy. Your private results stay private, and importing the request doesn’t start paid work.

## Work in projects.

Link collections and imports to a project, add a description or notes, and invite editors or viewers by email. Owners manage access and can change member roles and spending limits. Project and member limits apply to new work.

Add notes to a project or an individual record. Mention a collaborator with @email and resolve the note when it’s handled. Authors can edit their own notes; authors and project owners can remove them.

[Keep data current](https://specstream.ai/docs/schedules.md) · [Sources on every value](https://specstream.ai/docs/evidence.md) · [Desktop, phone, and browser apps](https://specstream.ai/docs/apps.md)