I think we can all agree that the term “hybrid integration,” as coined by industry analysts, is a transitional approach to the future state of systems and data integration, and not the end game. Hybrid anything (take hybrid vehicles, for example) typically represents the recognition that an emerging technology is not fully ready to stand on its own.
In the case of hybrid integration, the emerging technology (and driving factor) is the cloud. Organizations must find a way to incorporate cloud applications into their integration operations, while still accommodating on-premises applications. Hybrid integration platforms, which combine both legacy on-premises and emerging cloud integrations solutions in a bolt-on manner, are one way of accomplishing this.
For legacy integration vendors such as Informatica, TIBCO, webMethods (Software AG), Axway, etc. whose aging delivery models and core technologies weren’t designed to breathe the thin air of the cloud, the hybrid integration model is an opportunity to extend the relevance of their 15-year-old paradigms.
As a result, iPaaS, which in many ways has become synonymous with the cloud integration piece of the hybrid integration approach, is built upon the same flawed do-it-yourself, pre-built adapters approach as the legacy on-premises integration solutions currently failing us today. Yes, iPaaS vendors such as SnapLogic, Jitterbit, and Dell Boomi offer modernized technology stacks that band-aid the urgent need to bring cloud computing into the integration mix, but is that enough to leapfrog us into the future of integration?
I posit “no.”
As our CEO Bob Renner wrote in an earlier blog post, iPaaS misses the mark by focusing strictly on ‘productivity’ tools and hosting. In other words, iPaaS is just another iteration on an outdated integration model. For iPaaS to be sustainable, it needs to bridge to a fundamentally better approach that is capable of tackling increasingly complex integration challenges, challenges that not surprisingly have far less to do with the technology itself and far more to do with today’s business and data climates.
These challenges include:
- Ever-growing reliance on data to inform business initiatives and optimization. As data takes its rightful place as one of businesses’ most important assets, integration solutions must put data at the center of their operations, and not simply treat it as a byproduct of integration as is the case with iPaaS.
- Decomposition of monolithic enterprise applications and the resulting convergence of middleware (e.g. ESB, EAI, ETL, B2B, etc.). Most iPaaS solutions simply modernize the base technology and point it at the challenges of integrating cloud applications, which serves to further silo an organization’s integration operations. At a time when the distinct patterns of integration are blurring and we are seeing increasing overlaps in function, along with an increased need for real-time cohesion among various integration disciplines, we should be focused on transitioning away from siloed integration operations. iPaaS is simply incapable of doing this as it is an incomplete vision of the future state of integration.
- Increasingly sophisticated and demanding regulatory, compliance, and risk management environments. iPaaS and related concepts like citizen integrators and hybrid integration are by their very natures counter forces to the strict management requirements of PCI DSS, SOC2, HIPAA, 21 CFR Part 11, ISO, HiTrust, and so on. Rather than reduce compliance complexity for users, iPaaS simply moves the cost and complexity from the developers (those developing interfaces, connectors and integrations) to the CISO. In fact, iPaaS solutions mask this issue by stating that their solutions are compliant and secure, when in fact this is only valid before the first developer line of code or script is written, or the first user integration goes live on the iPaaS platform. These customizations, implementations and modifications are all “out of scope” for the iPaaS compliance certifications. In other words, just like software products, they are exceptionally compliant until you actually want to use them and then the compliance burden is yours and yours alone.
- Scarcity of IT skills, increasing compensation and retention considerations, and a competitive environment for specific technology “hot skills” (including integration developers). Any savings gained through iPaaS’ modernized productivity tools are largely offset by the higher costs of developers with “hot skills” in these new products. As companies work to hire, train, and retain these specialized resources, time to market with iPaaS integrations is slowed and the very real problem of resource scarcity is ultimately magnified, not eased. And if these solutions don’t “stick,” larger companies will never reach the competencies and efficiencies in aggregate that is required to drive down costs.
With its continued reliance on do-it-yourself integration and narrow focus on application integration, iPaaS is a bridge to exactly the same issues we have today with legacy vendors. In other words, iPaaS is a bridge to nowhere.
Luckily, there’s an alternate route. It’s called dPaaS (Data Platform as a Service). Like iPaaS, dPaaS features modern technology to help enterprises integrate ever-growing numbers of cloud applications and data sources. But dPaaS, through its managed services approach and unification of integration and data management capabilities, is much better suited to solving today’s (and tomorrow’s) true integration challenges.
(Want to learn more? Check out our dPaaS solution, called the Liaison ALLOY Platform.)
By Jonathan Razza, Sr. Director, Emerging Technologies