Uusi jäsen esittäytyy: Cluetail Oy
COSSin verkosto on laajentunut Cluetail Oy:n liittyessä mukaan yhteisöjäseneksi!
Cluetail Oy auttaa yrityksiä, organisaatioita ja henkilöbrändejä kasvamaan räätälöityjen liiketoimintatiedon, markkinoinnin, viestinnän ja kansainvälisen myynnin tukipalveluiden avulla.
Tutustu alta tarkemmin englanninkieliseen esittelyyn Cluetailin toiminnasta ja arvoista – miten avoimen teknologian edistäminen, yhteisöllinen kehittäminen sekä kestävä ja vastuullinen liiketoiminta näkyvät Cluetailin ratkaisuissa ja tekemisessä.
Cluetail joins the COSS ecosystem to advance information discovery
Cluetail became a member of COSS in January 2026 because we support the cause of open technology in general, we are actively involved in developing open source software, and we want to build a sustainable business within the ecosystem of these technologies.
At Cluetail, we are exploring how open, decoupled recommender technology can support public-interest use cases and commercial information discovery services alike.
Our long-standing interest in information flows
With our backgrounds in journalism, content marketing, project management, and international business services, we have always been interested in how information flows. For our own research and advisory work, we continuously look for better tools and processes to search, monitor, and discover reliable information efficiently.
When recommender systems were first introduced on social media platforms in the early 2000s, they seemed like the holy grail: a way to navigate growing information abundance intelligently. Over time, however, we have seen many of these systems decay into opaque, platform-controlled black boxes.
Algorithms as hidden gatekeepers
For many people today, recommendation algorithms play a decisive role in how they encounter news, analysis, and public debate. What we read, watch, and listen to is increasingly shaped not by editors or by our own deliberate choices, but by systems embedded inside large platforms.
Much attention has been paid to the effects of these systems, including filter bubbles, polarisation, and attention capture. Most recommendation algorithms are tightly coupled to the platforms that deploy them. They are proprietary, invisible, and optimised for business models that reward engagement rather than understanding.
From an open source and open internet perspective, this tight coupling is a structural problem.
Due to their business models, some large overseas platforms are incentivised to reinforce users’ biases and to polarise discussions. By amplifying extreme opinions, users tend to spend more time on these platforms, making them more “marketable” for advertising purposes. This dynamic is unhealthy for individuals and for society.
Recommendation as infrastructure
An alternative approach is to treat recommendation as infrastructure: a distinct layer that can be developed, audited, reused, and governed independently of any single platform or interface.
This idea is not new in the history of the web. Email, RSS, and the web itself succeeded precisely because core functionality was decoupled from specific services.
Recommendation, however, largely evolved in the opposite direction.
By decoupling recommendation logic from platforms, users can regain agency over their own content consumption and curation. They might choose, for example:
- • a purely reverse-chronological feed,
- • a thematic or topic-based filter,
- • or a discovery feed highlighting content shared by like-minded actors they were not previously aware of.
An open-source experiment
To explore this last option, we have been working with a team from the South-Eastern Finland University of Applied Sciences (Xamk) on an open-source recommendation system based on collaborative filtering.
It is designed to help people discover personalised information based on links shared by others with similar information consumption patterns. Its key characteristics are:
- • the recommendation logic is transparent and reproducible,
- • large language models are not part of the core algorithm,
- • there is no behavioural profiling for advertising,
- • and no data capture for resale.
Open infrastructure and sustainable services
Decoupling recommendation algorithms from platforms does not mean rejecting commercial activity. On the contrary, open infrastructure can enable more sustainable and trustworthy services.
At Cluetail, we are exploring this through Cluetail Discover: a commercial information monitoring and discovery service built on the same collaborative filtering principles. In this context, the value lies in the combination of technology, domain expertise, curation, interpretation, and client-specific insight.
During the development of our minimum-viable product, we initially tested the system using the theme of "open source" itself. This has given us a solid baseline set of sources and a growing understanding of the dynamics within that ecosystem. We are currently looking to help technology companies that need to stay informed about developments in open source. From there, we expect to either expand or further focus our thematic scopes based on recurring and emerging information needs.
Joining the COSS community
By joining COSS, Cluetail hopes to contribute to a broader conversation about how open source, open standards, and modular architectures can shape the future of information discovery. At the same time, we are keen to explore business opportunities together with partners in the open source ecosystem and to learn from others who are navigating similar tensions between public values, open source, and sustainable services.
For more information, please visit our website at cluetail.com, or reach out by email to Founding Partner Jos Schuurmans at jos.schuurmans@cluetail.com.