Never click through portals again
The AI crawling agent watches websites you specify and gets in touch as soon as your keywords appear there. Instead of going through tender portals, network sites and news sources yourself on a regular basis, you get reminded when something genuinely relevant shows up.
How the monitoring works
You specify two things yourself: the websites to be watched, and the keywords that matter to you. Those can be tender and procurement portals, network and association sites, news sources or the sites of partners and competitors.
The agent visits these sites regularly and checks new content against your keywords. If something matches, you get a reminder – with the find and its source, so you can look straight away.
The difference from a newsletter: you determine the sources and the keywords, not someone else's editorial team. And you do not get a bundled email with ten topics of which one concerns you, but a notice about exactly your topic.
Typical uses
Watch procurement portals for your services and regions – without searching manually every day.
Instead of collecting subscriptions, you watch the sources directly and filter what is relevant yourself.
Catch announcements, dates and funding calls from network sites as soon as they appear.
Follow announcements and publications from relevant market participants.
Finding tenders automatically instead of clicking through portals
The most common use is tender research. Anyone who wants to win public contracts has to go through procurement portals regularly – several of them, because federal, state, municipal and special-purpose bodies use different platforms. It is work nobody enjoys, and which therefore gets left until a deadline has passed.
With stored keywords the direction reverses: you no longer search, you get reminded. Whether this works for a particular platform depends on whether it is readable without signing in – portals behind a login are not straightforwardly accessible to an automated request.
The second use is market monitoring: publications by competitors, funding calls from associations, personnel news in trade media. The same applies here: you determine sources and keywords, not the editorial team of a newsletter whose selection you cannot influence.
Anyone looking for funding anyway will find a maintained overview of current programmes, with the official sources, on our page AI funding programmes (in German).
Setup and responsibility are yours: You set up sources, keywords and intervals in KOSMO yourselves – setup by us is not part of the scope. That means responsibility sits with you too: whether a site may be monitored depends on its terms of use and on copyright. Check that before adding a source.
Häufige Fragen
What can I have monitored?
Websites that are commercially relevant to you – for example tender and procurement portals, network and association sites, news sources or the sites of partners and competitors. Which sources you add is your decision – and with it the responsibility for whether the site in question permits automated evaluation.
Does this really replace newsletters?
For many use cases, yes. The difference: you determine sources and keywords yourself instead of adopting an editorial selection – and you get the notice on your topic rather than a bundled email.
How does the agent know what is relevant?
Through the keywords you store. That is how you control how narrowly or broadly the monitoring works, and you can adjust it at any time.
Do you set the agent up for us?
No. You create sources, keywords and notifications in KOSMO yourselves – the feature is built to be operated without us. Choosing the sources, and their admissibility, is therefore your responsibility.
What happens to the monitored content?
It is processed in your KOSMO instance – on-premise or in our German data centre. The same data protection rules apply as for the rest of your knowledge: no US cloud, no training on your content.
Does this work with portals behind a login?
Not straightforwardly. Sites that are readable only after signing in cannot simply be fetched automatically – and often the terms of use expressly forbid it. Freely accessible portals and sites are the normal case.
How often is it checked?
The agent visits the stored sites regularly; you set the interval to suit the topic. For tenders a tighter interval is worthwhile than for an association newsletter.
How is this different from Google Alerts?
You determine the sources yourself instead of depending on a search engine's index, and the finds land in the same environment as the rest of your knowledge – on-premise or in our German data centre rather than at a US service.
What else KOSMO can do
Groups control who sees which files.
Connect Nextcloud and OneDrive folders as a knowledge source.
Connect emails and send answers straight out.
Every answer with evidence – read it in the document itself.
KOSMO on your phone, with push notifications.
Induction, training and internal information in one place.
Best seen for yourself
In 30 minutes we'll show you what this feature looks like in your working day – with your own content.
