Monitoring Enterprise 2.0: Content First, Management Later

 Posted by at 1:36 pm  eCommerce, server
Sep 202009
 

by Akber Choudhry

In my experience with implementing Web 2.0 collaboration platforms for companies, this question invariably pops up:  What control processes do we need to put in place?

The answer to that  question is not that straightforward.  The extraction of knowledge and its accumulation and organisation has to be as easy and seamless as possible, while the commercial realities of running an organisation dictate that processes around content and its management be laid out first, and understood by all.

The solution is not hard: We are not trying to establish a new process, for if we were doing that, we would be defeating the purpose by asking busy knowledge workers to partake in a new process.  We are encouraging them to use common web-based collaboration tools to improve their collaboration.  No one knows where and how that collaboration will end up, for if  it was all known beforehand, there might not be the need for collaboration and knowledge management tools in the first place!  So, what do we do?  We monitor the collaboration and wait for patterns and styles to emerge and then manage them, while still not disallowing new patterns and techniques to emerge over time — eventually becoming a continuing process.


For example, a project or a community does their work and moves on.  Someone (a human or an automated process) is watching and monitoring activity according to a policy.  Emails are sent out before the collaboration is taken down and archived, or a human intervenes and changes the policy as it applies to that collaboration.  Activity directions may also be discussed with legal, HR and general management.  When risks arise, automated processes can be put in place to discourage the risky activity.  An example may be a customer-facing forum that may divulge a company’s intellectual property.

Enabling Web 2.0 in the enterprise brings multiple management tiers along.  The client is right there in the browser, which may be controlled by a portal, which again may be linked to a rules engine, that works on a persistence layer that talks to a database.  You get the picture!  This is yet another IT management hurdle as the many tiers, interfaces and services that come together to provide intuitive and productive user interfaces need to be managed in a predictable way, and their security and performance assured.  Again, the approach here should be to think of services rather than applications (SoA) and organise and govern non-critical services after deployment, and as the need arises, so that a realistic idea of their usage and security requirements are known.

Identity Management as the Foundation of Collaboration

 Posted by at 6:14 pm  eCommerce, server
Sep 132009
 

by Akber Choudhry

Identity analysis and structure is the cornerstone of any collaboration system. Some questions that need to be asked are:

  • Who are the people that are required to collaborate, and any possible groupings?
  • What is the level of their attachment to the host organisation, and which organisations do they belong to?
  • Do we need to verify actual identities, or are virtual identities good enough?
  • What is the mechanism of verifying their real or virtual identity?
  • What is the least obstructive authentication method for their role in the collaboration?
  • What amount of identity information needs to be stored?
  • What are the processes for provisioning and de-provisioning of entitlement, and what is the cost of these processes as opposed to their benefit?

Many other questions will be raised in specific situations, but the spectrum of collaboration activities (polls, surveys, blogs, forums, applications etc.) will build upon the identity management, authentication, authorisation and provisioning structures that have been set up keeping in mind the desired business objectives.