Products
CoJourno
Gambit 5
Newsroom Assistant
Stop archiving by hand.
"What ran on Tuesday's 6pm show?" "What's the gender breakdown of guests this month?" "Have we covered this angle before?" Your journalists ask these questions every day. The answers are buried across MAM, NRCS, rundowns and archive. Newsroom Assistant returns them in seconds.
What this Gambit does
Newsroom Assistant is an intelligence layer that sits on top of the systems your newsroom already runs - MAM, NRCS, rundown history, archive, guest databases, transcription archives. Your team asks questions in natural language. The Assistant pulls the answer from across the connected sources and returns it in seconds.
​
"What was the lead story on the 6pm show on Tuesday March 11?" Returns the rundown, the package sequence, the running order, the on-air talent.
"What's the gender breakdown of guests featured in this month's morning show?" Returns counts, percentages, and a comparison to previous months.
"Have we covered the new healthcare policy?" Returns relevant packages from the archive, with transcribed segments, dates aired, and which shows ran them.
"Which reporters covered the elections last year?" Returns the staff list, package counts, and links to the original packages.
​
The Assistant doesn't replace any system. It reads from them. The original tools - MAM, NRCS, archive - stay exactly where they are, and the journalists who prefer to use them directly can keep doing so. What the Assistant adds is a conversational layer for the questions that cross system boundaries.
The questions everyone asks, and nobody can answer quickly
Every newsroom runs on knowledge that is technically available but practically buried. Your archive holds the answer to whether you've covered this story before - but searching for it requires opening the MAM, knowing the right keywords, and trusting that the metadata was tagged consistently. Your rundown history holds the answer to what aired last Tuesday at 6pm - but reconstructing it means navigating the NRCS, finding the right show, and reading through the rundown line by line.
Your guest database holds the answer to whether you have a gender imbalance in your booking patterns - but compiling it requires someone to pull the guest list, cross-reference it with broadcast logs, count manually. By the time you have an answer, the editorial conversation that prompted the question has moved on.
Most of these questions don't get asked, because the cost of answering them is too high. The newsroom runs on intuition where it could be running on knowledge.
Where it sits
Newsroom Assistant connects to your MAM, NRCS, rundown history, archive, transcription store, and guest or talent databases via API. It doesn't replicate your data - it queries the original sources, returns the answers, and respects the permission structure each system already has. Journalists who can't see certain content in the MAM can't see it through the Assistant either.
The interface is conversational - accessible from the desktop, from the rundown editor, or as an in-context overlay in tools like Mimir and Saga.
What changes
The questions that used to take half an hour to answer take five seconds. Editorial decisions happen with knowledge instead of intuition. Patterns in your booking, your coverage, your guest mix become visible - not as a special research project, but as something an editor can ask about between bulletins. And the institutional memory of your newsroom - packages aired, stories covered, voices featured - stops being locked behind whichever system happens to hold it.
built by