Vijay Raina stands at the forefront of the enterprise SaaS evolution, bringing a wealth of knowledge in software design and architecture to an industry currently facing a profound identity crisis. As the “SaaSpocalypse” reshapes the digital landscape, Vijay offers a grounded perspective on how businesses can navigate the sudden surge of agentic coding and AI-driven disruption. His insights help bridge the gap between the raw excitement of new technology and the pragmatic requirements of large-scale enterprise stability, ensuring that organizations don’t just innovate, but do so with a sustainable foundation.
This conversation navigates the complexities of the agentic coding era, where the newfound ease of building custom software is forcing a massive re-evaluation of the classic build-versus-buy strategy. We examine the financial tremors that wiped $1.6 trillion from software stocks and why niche, low-context applications are most at risk of being replaced by “vibe-coded” internal tools. The discussion also touches on the shift toward outcome-based pricing models, the security risks posed by shadow AI in the workplace, and the essential strategies for SaaS vendors to remain relevant by offering a “service envelope” that extends far beyond a simple user interface.
How should a modern business weigh the immediate allure of “vibe coding” a custom solution against the long-term stability and accumulated experience offered by traditional off-the-shelf SaaS products?
The temptation to “vibe code” a solution specifically tailored to your company’s unique quirks is incredibly high right now because AI agents make the initial build feel almost effortless. However, I often caution leaders that standing up a piece of software is just the honeymoon phase; the real marriage begins during the years of maintenance that follow. In roughly 90% of cases involving administrative or line-of-business software, building your own version is actually a strategic mistake because you miss out on the “wisdom of the crowd.” SaaS vendors aggregate the needs and the hard-learned lessons of thousands of different companies, meaning their products are hardened by real-world use cases you haven’t even encountered yet. When you go the DIY route for core infrastructure, you aren’t just building a tool; you are taking on an unrelenting burden of patchwork fixes and technical debt that can eventually cause your team to drown. It is far better to reserve your “vibe coding” energy for customer-facing experiences or unique data products that provide a true competitive edge, rather than reinventing the wheel for back-office functions.
With software stocks plummeting by $1.6 trillion at the start of the year, what specific traits distinguish the vendors that are truly at risk of obsolescence from those that will remain indispensable?
The massive $1.6 trillion hit to the market was a wake-up call that “hollowed-out” SaaS companies—those that offer simple workflows with a nice UI but very little depth—are in serious trouble. We are seeing a new phenomenon where private equity firms and VCs are hiring services to “reverse build” existing SaaS offerings using agentic coding to see how easily they can be replicated. If a product can be cloned in a weekend by an AI agent, its valuation will naturally sink because the barrier to entry has vanished. The vendors that will thrive are those building “high-context” products that are encased in a larger service envelope, providing value that AI alone cannot replicate. These indispensable companies understand that the old formula of finding a workflow and slapping a price tag on it is dead; they must now offer discrete business process outsourcing or specialized services that make them a partner rather than just a utility.
As AI agents begin to handle tasks that were previously the domain of human employees, how do you see the transition from per-seat pricing to outcome-based models like pay-per-resolution changing the vendor-customer relationship?
The traditional pay-per-seat model is becoming an artifact of a pre-AI world because, in an era of automation, having more users isn’t necessarily the goal for a business. We are moving toward a value-based pricing structure where costs are directly aligned with actual outcomes, such as the model we see with Intercom’s Fin AI agent, which charges per resolution. This shift is exciting because it forces vendors to be more accountable; they only get paid when the software actually solves a problem, which perfectly aligns their incentives with those of the customer. However, this also introduces a new layer of volatility for vendors, as they must account for the variable, usage-driven costs of the AI agents running in the background. For the customer, it feels much more fair and transparent, turning software from a fixed overhead cost into a dynamic engine for ROI.
What are the hidden dangers of the “Bring Your Own AI” trend, and how can leadership teams balance the desire for employee productivity with the need for enterprise-grade security guardrails?
We are seeing a repeat of the early “Bring Your Own Device” era, where employees are bypassing official channels to use their own personal AI tools to “vibe code” assistants and automate their daily tasks. While this “shadow AI” can lead to a localized spike in productivity, it creates terrifying security risks because sensitive organizational data is being fed into external models without any guardrails. Most companies have already rolled out enterprise versions of Gemini or Copilot to provide a safe environment, yet the lure of familiar, personal tools remains a persistent challenge. Rather than trying to implement draconian rules that stifle innovation, leadership should focus on education and creating a path of least resistance for safe AI usage. The goal is to empower the “individual innovator” while ensuring that the data they are handling doesn’t end up in a public training set or a leaked database.
What is your forecast for the SaaS industry over the next few years?
I believe the next few years will be a period of intense creative destruction where the “SaaSpocalypse” actually clears the way for a more robust and diverse software ecosystem. We are going to see a “wild and potentially volatile ride” as businesses experiment with agentic coding, leading to some spectacular successes and many quiet, expensive failures as the reality of software maintenance sets in. The era of the “all-in-one” bloated platform may fade, replaced by high-context agents and specialized tools that are priced based on the actual value they deliver rather than the number of people clicking buttons. Ultimately, SaaS isn’t dying; it is simply shedding its outdated skin to become faster, more integrated, and far more aligned with the actual work that humans and agents do together. It is an incredibly exciting time to be in this space, provided you are willing to rethink everything you thought you knew about the cost to ship and the definition of a software product.
