Products
CoJourno
Gambit 4
AI Based Housekeeping & Archiving
Stop archiving by hand.
In every newsroom we've studied, a media manager's day is dominated by manual housekeeping. CoJourno's housekeeping Gambit turns that bottleneck into a background process - fully compatible with Mimir, with audit trails and human oversight built in.
What this Gambit does
The housekeeping Gambit applies rules-based automation with AI-driven inference. You define the policies - what counts as expired, what storage tier content belongs in, when working copies should be cleaned up, what metadata is required at which stage. The Gambit runs continuously against your MAM, identifying assets that match the rules and processing them without human intervention.
​
For metadata gaps, it uses AI inference: transcription where audio is present, face detection on speakers, object and scene recognition, semantic tagging based on the package script if one exists. Where confidence is high, the metadata is written. Where confidence is low, the asset is flagged for human review - not silently processed.
​
For movement between storage tiers, it executes the move and logs it. Working copies that are no longer needed get cleaned up. Assets that meet archiving criteria get archived, with the metadata enriched as they pass through the pipeline.
​
Everything is auditable. Every action the Gambit takes is logged with a reason - which rule triggered it, what the Gambit inferred, what was confirmed by AI vs. flagged for human review. Your library team can trace any asset's path through the housekeeping process.
Human oversight is the default for low-confidence cases. The Gambit asks for help when it's not sure, and learns from the answers.
The work that fills the day
In every major newsroom we've studied, a media manager - sometimes a whole team - spends the majority of their working day on manual archive housekeeping. Checking what's expired. Verifying that material has been properly archived. Deleting working copies that are no longer needed. Filling in metadata gaps before content moves to long-term storage. Resolving conflicts when an automated cataloguing tool flagged something the human can't quite agree with.
The work is essential. A MAM that fills up with orphaned media, missing metadata, and uncatalogued content grinds to a halt. Search becomes useless. Storage costs balloon. Archive integrity erodes.
But almost none of it requires human judgment. Most of it is rule-following. Some of it benefits from AI inference. A small fraction needs human review. Today, all of it is done by hand - because nobody had built the automation that respects how the work actually flows.
Where it sits
The housekeeping Gambit is fully compatible with Mimir and operates against your MAM via API. It doesn't replace your existing archive infrastructure - it sits on top of it, applying your policies continuously where today they're applied by hand on a periodic basis. It chains with your archive migration if you're running one. And the audit log integrates with your existing reporting tools so management has visibility into what's being processed and what's being escalated.
What changes
Media managers stop spending most of their day on manual housekeeping. Their time goes back to the work that actually requires editorial judgment - supporting journalists searching for archive footage, curating institutional material for re-use, managing rights and compliance. The archive stays clean. The metadata stays current. You stop paying for storage you don't need, and you stop losing footage you do.
built by