“We built monday in four hours using our system”
How AI-driven development is challenging the foundations of SaaS.
The world moves fast, but in 2026 it seems to have shifted to the speed of light. What was last year’s big hit, “vibe coding,” the practice of writing code using artificial intelligence, is already considered old news. The new industrial revolution is here, and it is not only changing how software is built; it threatens to wipe out billion-dollar business models.
Until recently, organizations faced a harsh dilemma: purchase rigid off-the-shelf software that rarely fit their exact needs, or build an expensive in-house development team that quickly became a bottleneck. No-code platforms such as monday.com and Salesforce, which made it possible to build applications without traditional programming, offered a partial solution, but they still tied users to predefined widgets and templates.
1 View gallery
Michal Lupo (from right), Amitay Gilboa, Oren Yunger, and Erez Shachar.
(Photos: Blocks, Spring.New, GGV, Qumra Capital)
Today, artificial intelligence is reshuffling the deck, allowing developers to create exactly what they need simply by describing it to a system that handles the technical work. “We built monday in four hours using our system,” says Amitay Gilboa, founder and CEO of Spring (spring.new). “In the new world, people with vision and creativity have more room than people with technological knowledge. The biggest barrier, development resources, has simply disappeared. Large language models already write better code than the average programmer in many organizations.”
Spring is a workspace that enables anyone to build business software using AI, simply by describing what they need, without writing code, without settling for almost-right off-the-shelf products, and without waiting months for development. “We are building for a world in which business software is not bought but created,” Gilboa adds, “and what differentiates companies is not the tools they pay for, but the ideas their teams are able to implement.”
The result is a democratization of production. In the past, a marketing manager might have begged the IT department to build an influencer-management tool; today, they can simply describe their needs to an AI system without opening tickets or involving other teams. The software is tailored like a glove, and built at a fraction of the cost of an expensive SaaS subscription.
The deepest change is not only in how software is built, but in what it does. Traditional software was passive: a repository that required humans to input data and manually operate interfaces. “Software is becoming active,” industry insiders explain. “An AI agent doesn’t just store information, it works for you.”
Imagine an agent that manages an entire marketing operation: it creates a campaign, runs it across social platforms, analyzes results in real time, learns what worked, and builds the next campaign on its own. In venture capital firms, similar agents are already scanning massive datasets and generating insights at a speed no human analyst can match.
Companies that currently spend tens or even hundreds of millions of dollars on software licenses are beginning to ask a simple question: why? With a single, agnostic platform that enables every operations manager to build their own tools, the SaaS model itself faces an existential threat. “The big companies are panicking, and rightly so,” says Gilboa. “The value of SaaS came from the fact that building software yourself was expensive and complex. Today, when you can build an entire vendor-management system in three days and cancel dozens of subscriptions to monday and similar tools, it threatens everything we thought we knew.”
Michal Lupo, co-founder and CEO of startup Blocks, agrees that platforms like monday helped teams build systems without traditional coding. “Before, companies either bought rigid off-the-shelf software or hired internal development teams, which quickly became bottlenecks,” she says. “AI now builds the code itself, allowing people to create simple, automated processes.”
Blocks enables users to build applications while receiving assistance from AI agents with advanced logic. Lupo describes the shift as fundamental: “In the past, people spent years learning how to write code. Today, you tell the AI what to build, and it does it for you. This opens programming to people without technical backgrounds. And unlike before, AI-generated code is highly flexible, the software can be built and designed exactly to fit the user. Software used to be passive; now it is becoming active. This is the next step in organizational efficiency.”
Yet the revolution also introduces new and complex risks. Oren Yunger, managing partner at Notable Capital, warns that traditional security defenses may soon become obsolete. “Malicious AI agents are already ‘smart-chaining’ vulnerabilities that were once considered insignificant,” he explains. “They create attack paths that slip beneath existing defenses. If humans were once the weakest link, today autonomous AI agents themselves have become the target. Flooding organizations with AI-generated code without proper engineering oversight also creates entirely new security risks.”
Despite the upheaval, not everyone is rushing to write the obituary of large cloud software companies. Erez Shachar, a partner at Qumra Capital, argues that the true power of companies like Salesforce and monday lies not in their code, but in their deep understanding of organizational processes. “Companies didn’t buy software, they bought knowledge,” he says. “Salesforce understands how sales organizations work. Their strength is in accumulated organizational data and domain expertise. The companies that survive will be those that are willing to disrupt themselves, stop defending legacy code and start delivering their knowledge through new AI-driven tools.”
