4 minute read
By Mike Bookey
Posted in Customer Engagement
It’s hard to imagine now, but a decade ago, open-source software—programs that allow users to modify the source code—was still on the fringes. Startups were starting to build on open source and open core, but few, if any, enterprises were.
Looking back, we can now say that open-source models undoubtedly accelerated both the pace of innovation and the quality of traditional software development. Nowadays, most anyone who is trying to build a successful SaaS product typically leverages as much open-source code as possible.
Given their success building open-source SaaS solutions, it makes sense that many enterprises would strongly consider building out their AI capabilities in house. The problem is that the enterprise’s experience with open-source SaaS isn’t a good corollary for open-source AI. While it’s true that any open-source project may come with some hidden costs, the hidden costs of open-source AI development are harder to anticipate and often severe.
Companies need to be aware of what they are getting into at the outset, especially as these decisions could have long-term implications across their organization.
It’s important to acknowledge that while previous successes with open source are pulling companies toward open-source AI solutions, the confused state of the flooded AI vendor market is pushing many in the same direction.
In their recently published Conversational Chatbot Buyer’s Guide, analyst firm Forrester notes that “the endless array of providers, the often indistinguishable service offerings, and the overuse of artificial intelligence (AI) and machine learning (ML) buzzwords” makes it difficult for businesses to define a clear and effective AI strategy.
For some companies, especially those that have very straightforward needs and a precise deployment strategy for single-point conversational bots, open source AI development may make sense.
One major consideration that Forrester points out, however, is that many companies are “tempted to use open-source tech to build chatbots instead of a market solution but fail to recognize the other financial and organizational support they will need to make this work.” Hiring a team of AI developers can cost up to $800,000 annually, if not more. Moreover, the crunch on AI developers across the market makes finding talent that can take advantage of even low-code or no-code options a major challenge.
And here at Verint, we’re hardly advocating for a walled-garden scenario where you’re locked into a specific backend system. In fact, Verint AI-powered automation solutions, including our conversational AI and intelligent virtual assistants (IVAs), are open and modular. So, you can start anywhere and truly go anywhere with enterprise-level capabilities and sophistication.
There’s also the reality that open-source options often prove challenging to scale. While initial deployments of conversational bots can address specific areas, many of these solutions fail to integrate into the overall infrastructure of an organization. Many don’t scale beyond the initial proof of concept.
Open-source solutions also prove popular for their perception of control. Organizations clearly want control over how their data is managed and to ensure that compliance and security requirements meet necessary standards. Here’s the thing, though—strictly in-house development can actually hamper this governance for a company, preventing them from leveraging the insights and data from AI point solutions and drive value across the entire breadth of their organization.
As a 2019 IDC survey found: “Most organizations reported some failures among their AI projects, with a quarter of them reporting up to 50% failure rate; lack of skilled staff and unrealistic expectations were identified as the top reasons for failure.”
These projects suffer from, among other things, a lack of pre-planning and data inspection in building the AI solution.
Vendor experience is invaluable when it comes to scaling AI and getting the most value out of your investment. They can share how to integrate AI solutions in ways that take them from mere precision tools to sources of upside insights that can meaningfully drive organizational strategy and change.
At Verint, our nearly two decades of developing AI solutions allow us to avoid many of the pitfalls created by open-source programs.
For example, Verint’s AI Blueprint provides the insight and planning that analyzes your company’s data to efficiently and effectively prepare your AI deployment for success from day one. AI Blueprint surfaces customer pain points and business opportunities through machine learning, and then creates a roadmap for long-term, sustainable, and scalable growth for your AI efforts.
A tool like this, and the expert service that comes along with it, isn’t available with an open-source AI solution.
Sure, open-source solutions and in-house development may be best for simple point solutions, but even as AI moves toward low-code offerings, scaling beyond the basics will prove much more difficult for most companies.
That said, Verint doesn’t hold your AI hostage.
Once we’ve established a company’s AI implementation, we can work with its employees to take the AI-powered system in house, should the company desire to do so. We’re here to help, even if that means giving you the tools to take the technology and run with it.
But remember, when it comes to AI implementation, it’s about finding the right experienced partner. This isn’t just to help companies achieve more robust AI integrations, but also to save them considerable time and money in the long run.
Did you like this story?
Subscribe for more Customer Engagement insights