Engineering depth: How tech companies can build to outlast the AI hype cycle

History repeats itself. The dot-com frenzy, the cloud revolution, the crypto gold rush, and now, AI. Each promised to reshape the world. Billions were poured in, fortunes were made, and then, just as quickly, the hype collapsed.

Each wave of innovation is accompanied by soaring expectations, wild investments, and an unshakeable belief in the success of this shiny new tech. But as one ‘big thing’ fades, another emerges—brighter and more hyped up than before.

Hype fuels urgency. Urgency breeds shortcuts. And in the rush to ride the AI wave, companies risk repeating the same mistakes—overpromising, over investing, and ultimately, overextending. So the real question isn’t who’s winning today—but who’s building to last?

The answer is simple: Companies with depth. Those who’re committed to engineering excellence, financial prudence, and know that real innovation isn’t about rushing to market but about building a foundation that can weather any storm.

But how to pursue depth while building a company amidst the pressure of market demands and an all-pervasive FOMO?

That’s exactly what Dheeraj Pandey, Co-Founder and CEO of DevRev and former CEO of Nutanix, and Amit Prakash, Co-Founder and CTO of ThoughtSpot, will zero in on in this episode of the Effortless Podcast. As industry veterans who have seen multiple hype cycles play out, from cloud computing to AI, they’ve built companies that thrived on engineering depth.

Why depth wins over hype in company-building

The tech industry is often characterized by its obsession with the new and the flashy. Companies pivot, rebrand, and chase after the next big trend, hoping to capture lightning in a bottle. But this lack of clarity and purpose can be a company’s downfall.

Amit, who worked at Google for close to 5 years, explains that Google was crystal clear about its identity as a technology company from the start.

Yahoo, in contrast, flailed between identities, unsure whether to be a media giant or a tech innovator. It spread itself too thin, chasing partnerships and acquisitions instead of doubling down on engineering. And in the end, that lack of conviction cost them the future.

This unwavering focus on engineering excellence meant Google could innovate from a position of strength rather than reactiveness. Google developed a robust technological backbone that allowed it to scale and adapt, Amit adds.

Amazon, too, had a similar commitment to depth right from its humble beginnings as an online bookstore. As Dheeraj points out, while many companies were content with outsourcing critical infrastructure, Amazon refused to trivialize essential components like identity management, billing, and metering. These weren’t mere IT problems—they were core engineering challenges that had to be solved in-house.

Unlike competitors that believed they could buy their way into scale, Amazon understood that depth in engineering was non-negotiable, which helped them build one of the most sophisticated, secure, and developer-friendly ecosystems in existence. Their depth in API design and security became a competitive moat, proving once again that true staying power isn’t just about reacting to trends but about going deeper than anyone else is willing to go.

Hype cycles and the financial reality of staying power

When the hype fades, the companies that survive aren’t always the flashiest, the fastest-growing, or the most well-funded. Instead, they tend to be the ones that understand the financial realities of building something sustainable, making disciplined trade-offs, and balancing long-term vision with short-term execution.

The dot-com bust of the early 2000s is a cautionary tale in financial discipline. Companies that relied solely on hype-driven funding collapsed overnight, while those with deep operational strategies and fiscal discipline, like Amazon, managed to navigate through the chaos and emerge stronger.

Balancing long-term vision and resource management

Many companies collapse because they fail to strike a balance between long-term vision and short-term decision-making. While vision gives a company its direction, effective resource management is what allows that vision to materialize. Without cash flow discipline, vision is simply wishful thinking.

Economic recessions, hype cycle corrections, and funding slowdowns are inevitable. Companies that fail to plan for these cycles will not survive them. The companies that remain standing are those that understand that capital is finite and trade-offs are essential, even in times of growth.

As Amit reminds founders and leaders, frugality isn’t about stifling innovation—it’s about ruthless prioritization. By investing only in what is essential, companies create a financial cushion that allows them to weather storms while still pursuing meaningful improvements in their core product offerings.

Depth and compounded learning

Companies that endure downturns amass experience, which becomes their competitive advantage. Their systems are engineered to be reliable, their operations are scalable, and they have repeat customers who trust them even during tough times.

Dheeraj’s point on companies compounding their learning speaks to the essence of staying power. Repeat business isn’t an accident, but the byproduct of reliable systems, disciplined operations, and a relentless focus on delivering value to customers.

The hype cycle and the trough of disillusionment

The Gartner Hype Cycle, which Dheeraj and Amit referred to earlier, is the reality check for every tech revolution. It maps out the rollercoaster of every tech innovation—the initial euphoria, inevitable crash, and eventual path to real impact—and AI is no exception to this.

Every groundbreaking technology follows this predictable 5-step trajectory:

• Innovation Trigger: It all starts with a breakthrough. A new technology—like AI, blockchain, or self-driving cars—sparks excitement. Investors flood in, startups emerge overnight, and everyone is convinced the world is about to change forever.
• Peak of Inflated Expectations: Hype takes over. Wild promises are made, like AI will replace industries, blockchain will end banks, and self-driving cars will make human drivers obsolete. FOMO surges, and billions pour into anything remotely tied to the trend, regardless of real-world feasibility.
• Trough of Disillusionment: Reality hits. The limitations become obvious. The tech isn’t mature, scaling is hard, and most companies can’t deliver on their promises. Funding dries up, weak players die off, and the industry faces a brutal reckoning.
• Slope of Enlightenment: The survivors start figuring things out. The technology isn’t magic, but it is useful—when applied correctly. Practical applications emerge, and companies that invested in deep, foundational work begin to see real traction.
• Plateau of Productivity: The hype is gone, but the impact is real. The technology is now widely adopted, deeply integrated, and delivering real business value.

Right now, AI is teetering between the Peak of Inflated Expectations and the Trough of Disillusionment. The companies that built for depth, engineered for reliability, and focused on real-world use cases will be the ones that survive the fall—and lead the future.

Depth is the key to reliable and resilient systems

Reliability is the silent foundation of every enduring technology company. Users may never notice when a system works flawlessly, but they will absolutely notice when it doesn’t. Whether it’s AI, search, or enterprise infrastructure, the difference between a tool people trust and one they abandon is often the system’s ability to be predictable, resilient, and correct—every single time.

As Amit recalls from his experience at ThoughtSpot, true depth in engineering isn’t just about designing for success—it’s about designing for failure. "Everything that we designed, we try to design it in a way that it’s a distributed system. If one node goes down, something else is there to take care of and the recovery is fast” he explains.

This kind of foresight is what separates robust, scalable systems from fragile ones. They were meticulously designed with redundancy, failover mechanisms, and self-healing architectures that ensured uptime even in the face of hardware failures, software bugs, or unexpected surges in demand.

Accuracy is a key element of reliability. Many startups optimize for speed first, and correctness later. But as Dheeraj points out, that approach is a recipe for failure, as this approach makes startups forget that returning customers matter.

This is especially relevant for AI companies today. Many are deploying unfinished, untested, and unreliable AI solutions, hoping that early traction will justify future fixes. But trust is fragile—if an AI system fails even once in a critical situation, users may never return.

LLMs aren’t enough—AI needs a structured operating system

In the early days of personal computing, software was fragmented, hardware-dependent, and unreliable. The emergence of operating systems like Windows, Linux, and MacOS provided a structured framework that abstracted complexity, standardized processes, and enabled software to scale globally.

AI is currently in that pre-OS era.

The current wave of AI startups is repeating the same mistakes that every past hype cycle has made—focusing too much on applications, not enough on foundational infrastructure.

Dheeraj’s observation about the lack of “operating system thinking” is crucial. The AI industry is currently building standalone applications without considering the underlying architecture required to support them at scale.

But instead of addressing these needs, many AI companies are focusing on building flashy applications without real infrastructure. They assume that large language models (LLMs) can function as retrieval engines, databases, and workflow orchestrators—all at once. But this is technically unsound and fundamentally inefficient.

Such fragile AI applications with “GPT wrappers” that rely on consultants, hacked-together pipelines, and limited automation will never survive, Dheeraj asserts. “It’s not product thinking, it’s not extensible, it’s not configurable. And it’s going to be extremely fragile,” he warns.

AI needs a structured foundation, not just a layer of hacked-together solutions. Without an OS-like approach, AI applications will remain brittle, limited, and unscalable. The companies that build robust, foundational AI operating systems will be the ones that define the industry.

According to Dheeraj, the OS-like foundation for AI must provide 3 things:

• Search capabilities: AI must be able to retrieve and process information efficiently.
• Workflow engines: AI agents must execute complex tasks beyond just generating text.

Analytics and observability: AI applications must be measurable, explainable, and debuggable.

Wrapping up: Depth is your survival strategy

In every hype cycle, there are two types of companies: those that chase trends and those that build the depth required to survive. The former burn fast and fade; the latter outlast the noise, weather downturns, and define the future.

Depth is what keeps startups alive. Not speed. Not hype. Not valuations. Depth in engineering, depth in financial discipline, and depth in learning.

The companies that make it through the hype aren’t the ones making the loudest announcements or raising the biggest funding rounds. They’re the ones that commit to the fundamentals—those that build a strong foundation while others scramble for short-term wins.

Many startups fail not because their ideas are bad, but because they don’t survive long enough to see them through. The market punishes those who sacrifice sustainability for speed, those who chase growth at any cost without building an operational backbone.

As Amit wrapped up his conversation with Dheeraj, he recalls the pithy advice that Dalton Caldwell, Managing Partner at Y Combinator, often gives startup founders: “Just don’t die.”

It sounds simple, but it’s the hardest thing to do in an industry obsessed with short-term wins. To sum up the key points from this conversation, the ability to outlast competitors requires a three-pronged strategy:

• Mastering cash flow and operational discipline so the company isn’t reliant on endless VC funding.
• Prioritizing reliability and engineering depth because a trustworthy product will always win over a flashy one.
• Making the right trade-offs to avoid cutting corners and ensuring every resource is spent on what truly matters.

The real lesson of every hype cycle is this: Survival is the strategy. Depth is the advantage.

When the hype fades, the market won’t remember who moved the fastest or who raised the most—it will remember who built the strongest foundation and delivered lasting value.

Startups don’t die because they lack good ideas. They die because they lack depth. And for the companies that take this path, the playbook is simple:

Just don’t die.

Want more insights on building for depth, not hype? Head over to The Effortless Podcast Substack. We break down first-principles thinking, hard-won lessons, and the realities of building enduring companies. No fluff—just the wisdom from two accomplished founders to outlast the hype.