After so much innovation and disruptive technologies, the issues surrounding legacy systems still prevail at scale.

For example, Global Banks are spending around US$220bn a year on IT expenses, with over 70% spent on maintaining legacy systems. These huge legacy maintenance costs are delivering less each year as the complexity of needs increases, whilst in parallel there is a gradual depletion of the knowledge required to maintain them effectively.

Let’s face facts! Legacy systems are so entrenched in organisations that they are here to stay until they are no longer required.

Generally, productivity at the macroeconomic level has remained stagnant for a long time, whilst digital disruption continues growing and growing.

However, digital disruption is rapidly leading to a stage whereby a tipping point is being reached that means legacy systems could be dealt with in a more innovative way.

Many legacy systems involve users performing tasks related to stakeholders such as citizens, customers, employees and suppliers around products and services. Many of these tasks are dependent upon knowledge, which are in other siloed systems such as explicit (documents, content etc) and tacit (in people’s heads). Its about time we looked at how the symmetry of knowledge and legacy systems work together in reality.

Knowledge-driven tasks are the ones, which are close to a tipping point being driven by digital disruption, which incidentally, have the potential biggest benefits to productivity and performance. Knowledge driven tasks are focused on the “last mile” involving the interaction and experience with the stakeholder (i.e. a citizen seeking planning permission; customer advice regarding a pensions transfer; the safeguarding of an employee; supplier guidelines for using their prescriptive drug).

Typically, both the legacy system and the knowledge are separate silos that co-exist not by intelligent design, but by necessity.  The perception is that these loosely coupled silos are orchestrated into a coherent workflow along with other siloed systems. Sadly, this perception is a chasm away from reality, especially when it involves complex knowledge such as regulatory, statutory, policies and procedures applied in practice. Though workflow provides coherence through bounded rationality and structure, the reality is that it is less able to cope with the increasing deviations leading to errors, falsehoods (false positives and false negatives), handoffs and in the worst-case negligence.  

Chatbots provide the opportunity to deliver conversation-as-a-service enabling a more intelligent encapsulation of legacy systems, whilst masking and overcoming their constraints. Chatbots enable the digital transformation of knowledge-driven tasks, with the benefit of compliance automation and emergent real-time evidence.

Whereas legacy systems are bounded rationality, chatbots by the very nature of conversations are not bounded and can rapidly evolve as needed. They have the additional benefit of being developed as small projects, avoiding the pitfalls of running large IT projects. More importantly, chatbots are co-developed with users and IT. Yes, users can create, share, measure and evolve chatbots.  

Behind the chatbot is the orchestration with the legacy systems, but more and more capability from the chatbots will overtime marginalise the dependency upon legacy. Chatbots do not end up replacing the legacy systems but are a catalyst to rethink legacy systems so the organisation is able to work more effectively and efficiently and adapt to every changing circumstance with its stakeholders.

The real issue regarding the siloism of knowledge and legacy systems is not the technology needed for digital transformation, but the development of new sense-making models, which sadly is still bounded by the normalisation of conventional thinking inside and outside the organisation. But, legacy systems are such a huge problem maybe those suffering the most will emerge as the new leaders for change.