Integrate And Surround Strategy

For IT and business leaders, the pressure to modernize aging systems has never been higher. However, full-scale replacements can be disruptive, risky, and costly. In the previous posts we talked about the modern monolith trend that helps companies simplify and streamline their application landscape while significantly reducing the operational costs. In this article we will explore a smarter modernization model: the Integrate and Surround strategy.

Few months ago I was researching on how to launch a successful Vertical SaaS offerings, and ran into this insightful article from Tidemark. As I read through it a familiar pattern started emerging. The methodology and approach PE / VC firms use to rapidly roll out Vertical SaaS platforms seemed to be taken right out of the hybrid Legacy Modernization playbook our team has been leveraging for years. I liked the term and we decided to adopt it.

• Implement a Meta Layer, typically a combination of API and Event driven middleware platforms. This is a critically important component and must be properly architected and rolled out from the start as it tends to bear the highest cost if done wrong
• Create modern UX apps that leverage legacy data and business logic provided by the Meta Layer, without a need to directly interact with / understand complexities of underlying legacy systems
• Deliver a modern Analytics platform that aggregates the data from various legacy systems and can run advanced analytics without impacting legacy systems operations
• Integrate other business systems with legacy platforms, typically done via Meta Layer to maximize the reuse and ensure proper system decoupling.
• Leverage AI and other disruptive technologies to further optimize the processes

• Meta Layer consists of API, orchestration, and data transformation components (we frequently work with Mulesoft and several other open source and commercial vendors), as well as Kafka event driven platform. The platform is used to surface the legacy functionality as APIs and integrate various systems into responsive “conversations”. The secret sauce for IBMi modernizations is InfoConnect product suite that turns any IBMi program into REST API, sends DB2 changes to Data platforms, and automates Green screen user actions, surfacing them as REST APIs, with little to no coding required.
• UX applications are modern front-end applications, built with React or Angular or other popular front-end stack, that leverage REST APIs provided by Meta Layer, and deliver the experience tailored to the business context and the interaction channel (mobile, web, CTI, kiosks, POS etc)
• Analytics is based on Azure Databricks / Data Lake Storage solution that ingests legacy data and business events, progresses the data through Bronze / Silver / Gold layers, and provides data visualization, dashboards, and reports via PowerBI and other reporting tools
• AI agents, chat bots, and AI-powered visualizations are built on top of Data layer and REST APIs and data definitions surfaced by Meta Layer

Integrate and Surround modernization strategy works well for variety of use cases, from isolated initiatives to large scale transformation projects, however there are several considerations and best practices. We listed the most important ones:

• Lack of integration strategy. Plan Meta layer with the target state in mind, earlier in the process, and treat it as a key enabling component. Companies often start with direct point to point integrations, and develop APIs directly on legacy platforms. While this may seem faster for the first couple of integrations, as the number of integrations grows, the overall solution becomes brittle, hard to manage, expensive to operate, and slow to evolve. Moreover, lack of governance and standards leads to potential security issues. By implementing battle-tested middleware solutions upfront, companies mitigate many of these downstream risks.
• Big bang or significant breadth of the initial scope. To make the project more manageable, start with just enough architecture, platform, and feature set that is quick to deliver and easy to tweak. Leverage the resulting prototypes to validate the technical solutions, platform choices, and intended user experience, without huge upfront budget and time investment. Once the small pilots succeed, the process can be quickly scaled out to cover other functional areas.
• Insufficient or missing measurable KPIs to monitor and refine the modernization project.
• Limited understanding of how the legacy system works. In order to get somewhere, we need to have a vision of the target state, the roadmap, but also, stating the obvious, know where we are now. Most successful modernization initiatives start with comprehensive analysis of most critical business processes, system operations, source code, data models, integrations, and analytics that the current systems support. Specialized AI-powered tools and modernization specialists with industry experience help significantly shorten this task.