AI Agents Enable Small Teams To Build Giant Things, Replacing SaaS

The era of the Small Giant: How AI is reshaping software development and the death of SaaS as we know it.

This conversation with Damian Tanner, founder of Pusher and now building Layercode, reveals a seismic shift in software development, moving beyond the industrialization of SaaS towards an era where small teams can achieve outsized impact. The implications are profound: traditional SaaS models, built around human-centric UIs and slow development cycles, are becoming obsolete as AI agents can now "do the work" more effectively. This transition necessitates a fundamental re-evaluation of development processes, team structures, and even the very nature of software interaction. Those who embrace this new paradigm, by learning to "trust the model" and leverage AI for rapid, bespoke software creation, will gain a significant competitive advantage. This analysis is crucial for founders, developers, and product leaders who need to navigate the obsolescence of established SaaS paradigms and harness the power of AI to build the next generation of software.

The Unraveling of SaaS: From Human Interfaces to Agentic Workflows

The software landscape is undergoing a transformation so fundamental that it signals the potential demise of the Software-as-a-Service (SaaS) model as we’ve known it. Damian Tanner, drawing from his experience building Pusher and now Layercode, argues that the era of SaaS, characterized by human-designed interfaces for human users, is drawing to a close. The advent of sophisticated AI agents capable of performing complex tasks, from coding to sales operations, renders the traditional SaaS UI increasingly redundant. Instead of learning to navigate a pre-built, often cumbersome SaaS tool, users can now simply articulate their needs, and AI agents can build or execute them.

This shift isn't merely about better interfaces; it's about a complete redefinition of how work gets done. Tanner recounts his own experience building a CRM for his company, Layercode. Frustrated by the complexity and slowness of existing SaaS CRM solutions, he used an AI agent to build a bespoke CRM tailored to his specific needs and preferences. This experience, he explains, was a "level one awakening" -- the realization that the software one needs can be generated on demand, rather than purchased and adapted. He subsequently canceled his existing CRM subscription, recognizing that the effort to learn and use the existing tool was far greater than the effort to have an AI build exactly what was needed. This suggests a future where bespoke software, created just-in-time and perfectly aligned with individual or team requirements, replaces the one-size-fits-all SaaS offering.

"The TLDR is SaaS, so there's a few layers. It's like the building of software or parts to software. There's the building of software and then there's the operating of software to get something wrong. And I think most developers are very familiar with like the building of software is changing now. But the, the operating of software, the operating of work, the doing of work in in all industries and all knowledge work can change. Like we've changed software and SaaS is made for humans, slow humans to use. The SaaS UI is made for a puny human to go in, understand, work out this complex thing and it has to be in a nice UI. If it's not a human actually doing the work that the, that that they do in the SaaS, if it's an AI doing that work, why, why is there a UI? Why is there a SaaS tool, right? The AI doesn't need a SaaS tool to get the work."

This perspective challenges the core value proposition of many SaaS companies. If an AI can instantly generate a functional CRM, a website builder, or any other tool based on a natural language prompt, the need for expensive, feature-rich, and often difficult-to-master SaaS applications diminishes. The advantage shifts to those who can effectively leverage AI to create custom solutions, bypassing the need for off-the-shelf software. This also implies a significant shift in competitive dynamics; companies that can quickly and effectively deploy AI to automate tasks or build internal tools will outpace those relying on traditional SaaS workflows.

The Demise of Code Review and the Rise of Agentic Development

The AI revolution is not only changing what software we use but also how we build it. Damian Tanner highlights a critical bottleneck in traditional software development: code review. As AI coding agents become increasingly capable, they generate code at a pace and volume that overwhelms conventional review processes. This leads to a scenario where code reviews become a significant impediment to progress, particularly for teams already struggling with AI-generated pull requests.

Tanner posits that the very processes designed to ensure human-written code quality--like code reviews, pull requests, and merge strategies--may become obsolete. These processes were established to mitigate human error. However, as AI models improve, their output becomes more reliable, potentially surpassing human capabilities in certain areas. The implication is that for new projects, where there's no legacy code to adhere to and no pre-existing "bad taste" in code architecture, AI-generated code might be trusted implicitly, bypassing the need for human review altogether.

"Processes that were built to ensure humans don't make mistakes, right? And help humans collaborate and help humans manage change in the repository and everything. It's like if the humans aren't writing the code anymore, we need to, we need to question these things."

This "trust the model" philosophy, while initially unsettling, is crucial for unlocking the velocity that AI agents offer. For new greenfield projects, teams can leverage AI to rapidly prototype and build entire applications. The focus then shifts from meticulous code review to prompt engineering and iterating on AI-generated solutions. This dramatically accelerates development cycles, allowing small teams to achieve the output previously associated with much larger organizations. The competitive advantage here lies in embracing this new workflow: teams that learn to effectively collaborate with AI agents, directing them to build and refine software, will be able to deliver features and products far faster than those clinging to traditional development methodologies. The "small giant" era, as Tanner calls it, is enabled by this radical increase in individual and small-team productivity.

The Small Giant Era: Amplifying Individual Impact

The convergence of advanced AI agents and cloud-native infrastructure is ushering in an era where small teams can achieve monumental feats, redefining what it means to be a "small giant." Damian Tanner emphasizes that the most significant change is not necessarily the technology itself, but the shift in mindset it enables. The ability of AI agents to rapidly generate code, automate complex tasks, and even build entire applications from simple prompts drastically amplifies the productivity of individual developers and small teams.

This shift liberates builders from the drudgery of repetitive tasks and the constraints of slow development cycles. Tanner encourages engineers to be more ambitious, to experiment with audacious ideas, and to "trust the model" even when it leads down unexpected paths. The low-risk nature of AI-assisted development means that even failed experiments provide valuable learning experiences, refining prompts and deepening understanding of AI capabilities. This iterative process, where code can be generated, tested, and discarded rapidly, mirrors the "molding clay" analogy, allowing for unprecedented exploration and innovation.

The infrastructure underpinning this new era is also evolving. Tanner's company, Layercode, leverages Cloudflare Workers, which provide global distribution, near-zero cold starts, and massive scalability without the need for traditional DevOps expertise. This allows small teams to build and deploy sophisticated, real-time applications like voice AI agents with remarkable efficiency. The focus shifts from managing infrastructure to orchestrating AI capabilities and focusing on the core problem being solved.

"I think the like the hottest thing is our own mindset, right? I I just found this with coding agents. You you start off putting in things where you kind of have an idea, you know what to expect out of it. And then you start just putting in stuff that just seems a bit ridiculous and ambitious. And oftentimes it fails. But more and more it's working. That's a very magical feeling. And and it's a very revealing kind of experience. And so I I think we can all be more ambitious now."

This democratization of capability means that individual developers, armed with AI agents and scalable cloud infrastructure, can now tackle projects that previously required large, specialized teams. The competitive advantage lies not just in adopting AI, but in cultivating an ambitious, experimental mindset that leverages these tools to their fullest potential. The "small giant" is no longer an anomaly but an emerging archetype, empowered to build and deploy at a scale previously unimaginable.

Key Action Items

• Embrace AI for Greenfield Development: For new projects, prioritize using AI agents (like Claude Code, Codex) to generate code and build entire applications from natural language prompts. This accelerates development significantly and allows for rapid iteration.
  • Immediate Action: Experiment with building a small, internal tool or a personal project using an AI coding assistant.
• Re-evaluate SaaS Dependencies: Identify SaaS tools that can be replaced by custom-built AI solutions. Focus on articulating needs clearly to AI agents rather than adapting to existing software limitations.
  • Immediate Action: List 3-5 core SaaS tools your team relies on and explore if AI can replicate or improve their functionality for your specific use case.
• Develop Prompt Engineering Skills: Invest time in learning how to craft effective prompts that elicit desired code and functionality from AI agents. This is the new core skill for developers.
  • Immediate Action: Dedicate 1-2 hours per week to practicing prompt engineering for code generation and task automation.
• Experiment with Voice AI Integration: Explore how voice interfaces can enhance existing applications or create new user experiences, leveraging platforms like Layercode.
  • Over the next quarter: Integrate a voice agent into a demo application or internal tool to understand its potential.
• Adopt a "Trust the Model" Mindset (for new projects): For greenfield development, be willing to trust AI-generated code and iterate based on its output, rather than immediately defaulting to traditional code review processes.
  • This pays off in 3-6 months by dramatically increasing development velocity.
• Focus on Mindset and Ambition: Cultivate a mindset of ambition and experimentation. Recognize that AI tools enable individuals and small teams to tackle larger, more complex problems than ever before.
  • Longer-term investment: Encourage a culture of experimentation and learning within your team, where trying ambitious AI-driven projects is valued.
• Explore Scalable, Serverless Infrastructure: Investigate platforms like Cloudflare Workers for building applications that require high scalability and real-time capabilities without heavy DevOps overhead.
  • Over the next 6-12 months: Evaluate migrating a non-critical service or building a new feature on a serverless platform to understand its benefits.