Request demo now

Connected Data London Recap


Conference co-organizer George Anadiotis invited me to share more about CELUM’s graph-based platform at Connected Data London 2018 after stumbling upon one of my technical blog articles. Connected Data London brings together experts from the fields AI, semantic technologies, linked data as well as graph databases.

Specifically, I told the story of how we built a production-ready, graph-based enterprise application in the cloud.

This application was our Agile Teamwork Management solution Workrooms. Needless to say that we used Workrooms to organize our conference attendance:

My talk demonstrated the technical advantages of combining a well-defined graph data model with the principle of reactivity. In a nutshell, our platform:

  • provides users with a snappier user experience through reactivity – the most relevant data is retrieved first which makes applications more responsive even with growing data sets or poor connectivity. Imagine reviewing a slide deck via Workrooms under time pressure on an airplane – our platform will provide you the essential data about your slide deck first so that you do not have to wait for loading auxiliary information like the version history that is irrelevant for your current task. This is possible by combining the graph principle of linking related information with the concept of reactivity.
  • allows software engineers to become more efficient by using a data model that feels more natural than traditional approaches and is less error prone due to a sophisticated schema
  • will allow users to model their own environment more naturally in the future as they are the experts on how their assets relate to each other

As we believe that these principles are applicable to a wide range of products, we decided to share the findings of our journey. This gives our listeners an edge when building their own graph-based application: 

Speaking of gaining an edge, Connected Data London provided three conference tracks with different focus:

  • Nodes – featuring more established topics
  • Edges – showcasing cutting edge approaches such as our platform
  • Educational – diving interactively into a topic

We hopped through the tracks and mingled with a diverse set of people from different sectors – leaders, researchers and practitioners. Specifically, I enjoyed engaging in deep discussions about graph database architectures with the attending graph database vendors. The small size and narrow focus of the conference really allowed to have valuable conversations.

Finally, our key takeaways from the conference are:

  • Graph technology is still an emerging field that leaves room for innovations and lively panel discussions. Specifically, Markus Lanthaler’s keynote on JSON-LD showed that it is important to start with simpler features that provide fast adaption and prove their usefulness rather than starting with too high ambitions that raise barriers for adoption.
  • Graphs are the basis for interconnecting diverse data sources to uncover new knowledge and thus, provide more value to the users
  • Yet, with these diverse data sources, issues of trustworthiness arise, as we need to extract the truth from conflicting information. According to the panellists, artificial intelligence could identify the truth via graph analysis.
  • All roads lead to Rome: Both rule-based AI as well as machine learning add value when extracting new knowledge from graphs.
  • Specifically, I was intrigued when Thorsten Liebig showed how humans and software complement each other: The software provides different representations of the graph data. The user contributes to knowledge discovery by identifying patterns from the data. Then, the user codifies the new knowledge and the machine derives further semantically correct conclusions by analysing the graph. Finally, the cycle repeats.
  • We at CELUM do existing stuff that matters so that our users can enjoy the benefits of graph technology