Every image geolocation tool has the same interaction model. Upload an image. Get a pin. Close the tab.
That is not how real work happens. Investigations involve dozens of images. Evidence accumulates across sessions. A location confirmed in one frame opens questions in three others.[3] The work is not a single query. The work is a process.[1]
Oceanir Desktop was built for that process. Not for the upload. For everything that comes before it, after it, and around it.
A new category
The easiest way to describe Oceanir Desktop is by what it is not. It is not an AI geolocation app. Not a reverse image search. Not a vision model. Not an OSINT tool.
Those are features. They describe components of a capability. They do not describe what you do with the capability.
The right category is Visual Intelligence Workspace. That category implies investigations, workflows, organization, evidence management, and professional tooling. It implies that serious work happens here, not just queries.
The distinction matters because it changes who the product is for, how it should be evaluated, and what replaces it when something better comes along.
Where it fits
Every discipline has a workspace that became the default surface for serious practitioners. That workspace is not just a feature. It is the place where the discipline's artifacts live.[2]
The workflow
Most tools answer one question. Desktop is built around the full sequence of work.
01
Ingest
Drop a folder of images or videos. Desktop reads them natively, no upload required.
02
Organize
Group evidence into cases. Each case holds files, analyses, notes, and tags in one place.
03
Geolocate
Run Orca on any piece of evidence directly from the workspace. Results attach to the case automatically.
04
Compare
Side by side candidates, contradictions, and confidence levels across the entire evidence set.
05
Annotate
Add notes and tags to individual files. Build the record as you work, not after.
06
Investigate
Follow a thread across images. Cluster by location, time, or visual similarity.
Why workspaces win
Models commoditize. The accuracy gap between competitors narrows over time. Any precision advantage is temporary. What does not commoditize is the workflow built around the output.
When Figma replaced Sketch, it was not because Figma had better vector rendering. It was because Figma made the artifacts live in a shared, persistent, organized workspace instead of someone's local filesystem. The model changed. The business changed. The stickiness is the workspace, not the output.
The same principle applies here. The prediction is the commodity. The workflow around the prediction is the product.
The value is not just the prediction. It is the workflow around the prediction.
Why desktop, not a browser tab
Investigations do not respect browser session boundaries. A case that spans three days, fifty images, and four analysts cannot live in a tab. It needs persistent state, local file access, and an interface that behaves like a real application.
Desktop installs as a native application. It reads files directly from disk. Evidence never has to be uploaded to a server to be processed. Cases persist across sessions. The workspace stays where you left it.
This is not a web app with a different frame around it. It is a different architecture for a different class of work.
Available now, in development
Oceanir Desktop is available as an early release. The core investigation loop is working: ingest, organize into cases, run analysis, and review results inside the workspace.
The features being built toward next are multi-file compare, evidence clustering, and export. Each one is part of the same thesis: the work happens in the workspace, not outside it.
