The complete desktop package for data work.
Agents, pipelines, playbooks and memory, running on the Mac in front of you. Point it at a folder or a warehouse and work the way you would in a spreadsheet — except the spreadsheet does the work.
Runs on your laptop.
- Data stays on this Mac
- Connected through your VPN
- Private by default
Why the data work belongs on your Mac
Most of what a data team handles is not allowed to leave the building. The desktop app keeps it there, and still gives you the whole harness.
Private by default
Files and query results stay on disk. Nothing is uploaded unless you say so.
Works behind your VPN
Connections start from your Mac, so warehouses on the private network just work.
Folders with permissions
Add a folder and the agent asks before each read, or always, per folder.
Every format you already have
Excel, Word, PowerPoint, PDF, images and CSV in; tables and files out.
The same harness as the cloud
Agents, pipelines, playbooks, lens and memory — not a lite edition.
Burst to the cloud when you choose
Send a heavy job or a schedule to your workspace. The raw data can stay put.
Local, hybrid or cloud — per project
Where a project runs is a setting, not an architecture decision. Change it later without rebuilding a pipeline or a playbook.
Local
Everything runs on the laptop. Nothing reaches Corvic's servers.
- Works with no outbound access to Corvic
- Bring your own model keys, or a model you host
- Results stay on disk
Hybrid
Raw data stays local. Jobs you approve run in your workspace.
- Heavy pipelines and schedules burst to the cloud
- Share dashboards without sharing source files
- Approve what leaves, per job
Cloud
A native front end to your Corvic workspace, shared with the team.
- Always-on pipelines and agents
- Private connectors to your warehouse
- SOC 2 compliant, same controls as corvic.ai
From download to first answer
- 1
Download and open
Sign in with your Corvic account, or start without one and stay local.
- 2
Add a folder or a source
Point it at a folder on the Mac, or at Snowflake, Databricks or Oracle.
- 3
Ask, build, schedule
Ask a question, turn the answer into a pipeline, and put it on a schedule.