Content Copyright © 2014 Bloor. All Rights Reserved.
Also posted on: The IM Blog
Diaku (which is from the Greek meaning to govern) develops and markets a product called Axon (as in the connectors between brain cells), which is difficult to classify. The easiest classification for the product would be to say that it provides a collaborative environment for data governance. However, while that is true it is not the whole truth. The company actually describes Axon as providing “collaborative enterprise understanding”, which is also true but isn’t an identified market yet.
Let’s start with data governance. Axon does not do data profiling, data matching or any of the other things that data quality providers offer: it integrates with these systems (and with script-based solutions) but is more concerned with understanding data and its interconnectedness both in terms of data flow and the business context it operates in, workflow, assignment of tasks and so on. In other words, the policy, understanding and collaboration side of data governance.
Actually, that’s not quite true. The one thing that Axon provides and that data profiling tools also provide, is dashboards that are used to monitor data quality. However, whereas a data profiling tool would simply measure quality at a table level, Axon is typically used to monitor quality with respect at a business level, whether that be for processes, projects, policies and so forth. In particular, there is an add-on module called Axon Regulation Insight that is designed to provide regulatory support. For example, suppose that you are implementing a KYC (Know Your Client) project, then you would likely build a dashboard that monitored all the relevant data that was specific to that project. Similarly, you might have different dashboards specifically related to the Volcker Rule or FATCA or BCBS239 (part of the Basel regulations) or anti-money laundering, and so on.
The actual way that Axon works is in three stages. The first stage is to create an inventory of relevant resources (which may include spreadsheets, text [even paper forms], Access databases and so forth, as well as relational and other databases). Next, you collaboratively and progressively describe what is in those sources and map the connections between these sources—for example where references to clients appear in multiple places. And finally, you allow business people (this is a very business-centric product) to explore those connections using their own terminology (Axon includes a business glossary that has machine learning capabilities so that it improves as time goes by).
Under the covers, Axon is based on the open source OrientDB, which is a document-graph database. That is to say it is a document database (hence the support for text and spreadsheets as well as relational data) that uses the concept of edges to connect associated sources and which allows Axon users to explore relationships (lineage, dependencies and so on) at scale. While OrientDB includes a SQL layer, the support for edges effectively means that you use these to traverse relationships rather than employing a JOIN statement, which makes traversal very much faster.
Anyway, this is not a discussion about OrientDB. Perhaps the key point that I have not mentioned is that Axon supports governance of not just the data that is relevant to any particular project but also the processes that are involved in implementing that project. And all stakeholders are collaboratively involved in that governance. Which brings me back to OrientDB: when you have multiple processes and multiple people undertaking multiple tasks against multiple data sources, then doing things like impact analysis (which Axon supports) can get quite complex. Without a graph capability of some kind it can easily become too difficult and slow to explore the environment.
To conclude: this is early days for Axon. It has primarily been working in the banking sector though it has recently acquired its first client in the pharmaceutical industry. It is clear that Axon is especially well-suited to environments that are heavily regulated. However, it clearly has wider applicability to data governance in general as well as the putative ‘enterprise understanding’ mentioned above. I think Axon is seriously cool: it’s definitely worth a look.