Can Agile work for Infrastructure and Ops Groups?

Yes. Agile can work with Infra and ops groups. This article outlines some basic principles during your agile transformation journey.

When an organization gets more and more Agile soon they realize that one of their bottlenecks in their own IT operations department. With the concept of physical datacenters slowly moving to the cloud the nature of work done in operations groups are also changing drastically with time.

The DevOps movement has been successful in moving parts of the operations, upstream so that the engineering teams can mostly be self sufficient and not have to deal with handovers and delays due to large wait periods before things get deployed. Concepts like SAAS, IAAS, PAAS, Cloud etc fundamentally changed the notion of network engineers, DBA and many roles that were typically there in classical IT operations groups.

Organizations that have tried Agile in Operations are moving from processes like ITIL and are heavily invested in specialized skills, network engineers, DBA’s etc.

Here are some pointers to transforming your Infra and Ops group to an agile organization.

First things first

  1. Identify all the Service Lines that your group supports and align cross-functional scrum teams around these service lines. For large services, consider using the concept of PODS which compose many small cross-functional feature teams. Set up Scrum of Scrums across PODS. These teams can follow a Scrum or Scrumban model of working. Use dedicated Scrum master for a team or an agile coach for a POD model . Teams need help during the change. Don’t ignore this role.
  2. Identify one or more product owner for the service lines.
  3. Move some of the operations folks upstream to directly work with engineering teams to enable DevOps and provide the necessary automation required to seamlessly deploy to production.
  4. For incidence response teams, involve the actual dev teams in the process. The first level response teams may not be able to use Scrum as most of the work they do may be demand based.
  5. Streamline ticketing systems so ticket duplications in systems like Service Now are removed. The same ticket should be used by teams to solve the incident. In earlier organizations since the org was siloed each group had its own ticket ending up in a lot of duplication and coordination of work.
  6. Education for leadership and middle management is key.
  7. Consider rolling this change to the entire org in some sort of structured manner using internal and external coaches so that there is minimal disruption to the operations.
  8. Set up leadership impediment board to relentlessly remove impediment across the organization.
  9. performance measurement systems should be changed to encourage more cross-functional learning than a siloed expert based learning the culture.