Skip to content

Tag: architecture principles

Digital transformation need for high flexibility and short time to market has been conflicting for several years now with robustness and latency of evolution of legacy systems. Bi-modal has been highlighted as being the silver bullet. Unfortunately most bi-modal IT systems are now only not maintainable mess. However, successful patterns exist.
Published on Leave a comment on Bi-modal IT is in essence a failure for digital transformation

Organizations and IT systems need to switch Bee-modal to cope with digital transformation

Facing Reality

We all know the great challenges introduced by digital transformation : Business models are evolving at unprecedented pace, competitors appear proposing disrupting solutions and partnership becomes vital. Convergence of exponential IT capacities evolutions (Big data, IoT, Artificial Intelligence, Cloud…) and structuring underlying society mutations (millennials, retired…) is bringing new customer needs and behaviors. These challenges are also wide areas of opportunities.

Unfortunately, past heavy investments on core IT systems have led to efficient but far from nimble structures. Legacy systems are now perceived as a major business development show-stopper.

Here appears bi-modal IT. Park legacy aside and work on next-gen IT. This can only be a temporary solution as it enables moving forward but without fixing the core issue. Legacy is still there and new systems are piling up in an anarchic way. The wait and see strategy  has long lived but is not sustainable anymore.

Thinking about it, isn’t data the new fuel for business growth and innovation? Here comes Data Centric Architecture, the key to unlocking this hidden value. It can even help to integrate or better use legacy systems by breaking data silos. Certainly managing and analyzing data can give real headache to Information Systems Directors but does not solve enterprise expertise strengthening or quest for innovation. Choosing the appropriate investment and pattern within all these Service Oriented Architecture, eXperience centric Architecture, Process Driven Architecture, Intelligent System Architecture is a blind bet. Which one is the closest to being “Digital Ready”?

Every single one is viewing through a zero sum lens. There exists nothing such as “Comprehensive Digital Architecture”. There are multiple paths and solutions to overcome digital challenges. All these rely on common foundations and guidelines. The premise is to shift from Bi-Modal IT organization to Bee-Modal.

Hive IS architecture for digital transformation : "Harmonized, Interactive, Variable, Elastic"
Digital IS overall architecture principles

HIVE IS digital architecture principles for digital transformation

Bees can build highly scalable and robust structures based on one single pattern and genetic understanding that working together makes the hive cleverer than being alone. Same goes for organizations and in the end IT systems.

These do not intend to be a single pattern or end target, but experience based guidelines to define the overall direction the organization and IT system need to follow in order to benefit from existing IS strengths, cope with digital transformation today and  anticipate future evolutions.

Harmonized 

The first character of a successful organization is its ability to utilize full potential of all its resources. All business lines are targeting a same goal and can even share this goal via extended partnerships. This of course does not mean bringing anarchic individual initiatives.

  • Foundation values and principles need to be formalized
  • Intrinsic strength and expertise need to be understood and confirmed
  • Based on these inputs, an overall investment and transformation strategy needs to be put in place balancing time to market and property, commodity and business differentiators
  • Appropriate partnerships need to be identified consistently with this defined build/buy/ally strategy
  • Strategy, goals and boundaries need to be clearly defined jointly by all stakeholders
  • Observations, facts, knowledge need to be consolidated and common standards built (e.g. : business definitions)
  • They need to be analyzed (individually or centrally)
  • And they need to be shared for better global efficiency

Interactive 

User experience is often considered as customer’s online experience. And buzz words like “mobile first” or “ATAWAD” (AnyTime, AnyWhere, AnyDevice) are the guidelines. This acronym is often wrongly interpreted (Responsive technologies and budget constraints help having an oversimplified vision).  Of course an end user should be able to access services when needed and where needed. But are you doing the same thing on your mobile with no network in the subway, on your tablet at home or on your computer at work? I don’t. Services need to be adapted to the context.

  • Interactions become proteiform. Push notifications replace mobile app consultation, chats replace Frequently Asked Questions and smart device become entry point instead of computers or mobiles.
  • User experience is not limited to customer experience. Employees and partners shall benefit from the same consideration. Partners need to be fully on boarded by facilitating exchanges and business users need to be given freedom of interaction with IS systems and ease of evolution.
  • IT system need to be open to the overall ecosystem, whether talking about virtual interactions (open data, web usage…) or physical ones (Internet of Things)
  • Human is kept in the center. As an example, regarding connected health initiatives, if quantified-self devices are mandatory, they cannot replace human contact and coaching.

Variable 

Overall application lifecycle need to be shortened. Context is indeed rapidly changing and fail-cheap or test-and-learn approaches are no longer an option.

This has a direct impact on the system:

  • Of course time to market needs to be reduced and application lifecycle fully equipped in order to enable continuous integration and DevOps. This shall be enabled with hybrid target in mind, whether using on-premise/cloud mix or heterogeneous cloud solutions (multiple providers, multiple hosting solutions)
  • They shall be compliant with Agile approach. This does not mean, though, that all projects and all components are eligible to Agile. Indeed, balance between agility and commitment to results needs to be measured and driven.
  • Beyond pure IT capabilities, reducing time to market or project complexity is brought by more and more efficient collaboration between teams.

Effectively, change is almost impossible without industry-wide collaboration, cooperation and consensus.

Simon Mainwaring

Elastic

Anticipating use of a new service, or paying for on-premise servers scarcely used are not a constraint anymore.  In addition, powerful new capabilities are brought like machine learning or artificial intelligence suite. 

Thanks to cloud capabilities, it is made possible.

  • A strategy on cloud usage shall be defined and refined for each service. If a cloud based service can be interesting for an emerging service, it can kill the RoI for high volume of activity.
  • Externalizing features, becoming a broker, reinforces the need for ease of connection and security. IS goal is not anymore to build a high fence around IT, but to ensure efficient point to point connections. This is a major change to security paradigm shift for CISOs. This is also to be considered for human resources management. As an example, is it worth hiring data scientists?

Going forward

Take into consideration that identifying appropriate components, orchestrating them and progressively putting in place foundations that stimulates business and not block it, is a vital challenge and can only be achieved by setting in parallel appropriate governance. What would a big data platform be without Chief Data Officer and Data Steward?

Feel free to share your own experiences. Did you experience long term success getting rid of (or coping with) legacy systems?