Remyral
08Research and documents · Researchexample setup · demo data

A sourced market report on your desk every Monday.

An example setup for a team tracking a market, competitors or regulation that wants reports and tables, not a repository: research becomes recurring work.

Start a pilotHow it connects

It runs on your own server: work files and records are stored on your infrastructure; content processed by AI goes to the model provider you choose, and data sent to external tools goes to the connections you set up.

demo data
01Workflowfrom the chat or the editor

Weekly market report

A chat request becomes an automation: the agent reads the web, compares with the previous report, and the report passes a person’s check before it reaches the team.

demo data

Last run · Mon 07:00 · 23 sources

02Setup and flowResearch and reports

Setup and flow

A team tracking a market, competitors or regulation that wants reports and tables, not a repository.

Setup

Project type
The built-in General workspace or a workspace project of your own; PDFs, CSVs and notes can be added as input.
Network
The open web (HTTPS only) or an allow-list of up to 64 addresses; local network and metadata addresses always blocked, every network decision logged.
Machine
A worker that supports workspaces, assigned to the project, and a runtime account.

What a day looks like

  1. 01

    You type “e-scooter sharing market: players, prices, news from the last three months” in the chat; the assistant drafts a research job in the General workspace and you approve it. The agent on the machine does the work.

  2. 02

    The agent reads the pages its network access allows and asks you through a question card when the scope is unclear; it writes a report.md with a source list and a players.csv to the delivery folder.

  3. 03

    The report and the table are previewed or downloaded in the Outputs screen. Turn the same job into an automation and every Monday brings a version updated on top of the previous report.

What changes

Research becomes recurring work that builds on the previous output; the report and the table sit in one place for the whole team.

How it connects

The agent reads the open web, allow-listed addresses or the files you add. If you use a paid search or data service, you connect it through an MCP server or the workflow HTTP step and set it up your way.

We don’t impose a ready-made flow; design your own, either with AI (describe it in the chat) or yourself (in the workflow editor).

Good to know

  • A team member does the final read of the report; the source list makes that easy.
  • For sensitive topics you keep the network on an allow-list and keep secrets out of prompts and files.
demo dataA question card in the chat: the agent asks a multiple-choice question where it cannot decide.
A screen from the console · The real console: a question card, when the scope is unclear the agent asks and you choose.
03Other use cases

Another team’s recurring work

All use cases and the connection process

Let’s fit this setup to your own work.

This example runs on the same infrastructure as all the others: Remyral on your own server, and we build your flow together in the pilot. Just describe the work you want to simplify and the tools you use in your application.

Start a pilotInstall and upgrade