Presenters

Source

Building the Future: How Retool is Navigating the AI Revolution and Redefining Internal Tools 🚀

In the fast-paced world of tech, staying ahead means not just reacting to change, but actively shaping it. David Hsu, CEO of Retool, joins us to share his insights on how his company is not only adapting to the AI revolution but also leading the charge in transforming how businesses build and operate. From strategic pivots to the nitty-gritty of product development, David offers a candid look at the challenges and triumphs of building a company in the age of AI.

The Builder’s Mindset: From Strategy to Code 👨‍💻

David Hsu emphasizes that building at Retool extends beyond just product development. It’s about cultivating the company itself as a dynamic, evolving entity. As a builder at heart, David actively participates in the creation process, even experimenting with cutting-edge AI tools like ClaudeBot (now Boltbox). This hands-on approach fuels Retool’s core mission: empowering users to build.

  • Retool’s 10-Year Journey: The company is gearing up for its 10th anniversary, a testament to its longevity and evolution in the tech landscape.

The AI Inflection Point: Shifting Focus and Embracing New Users 💡

Retool’s journey began in 2017 with a mission to simplify internal tool development for engineers. However, over the last 12-18 months, a significant shift has occurred. The rise of AI, particularly Large Language Models (LLMs), has accelerated a trend that started even before AI’s mainstream adoption: a surge in non-engineers building sophisticated applications on Retool.

  • From Engineers to Everyone: Initially catering to engineers, Retool now sees a majority of its builders as non-developers who can leverage tools like SQL to create full-fledged production applications.
  • AI as an Accelerator: While the trend of non-technical users building started pre-AI, LLMs have significantly amplified this movement, pushing Retool to “go all in” on this broader user base.

Navigating the AI Landscape: Bold Decisions and Hard Lessons Learned 🎯

The proliferation of LLMs presented Retool with a critical strategic question: how much to lean into this new technology while safeguarding their existing, successful business? David shares a nuanced perspective, highlighting both successes and missteps.

  • The “Startup within a Startup” Success: Retool launched “Retool Agents” in May of last year. This product, focused on using LLMs for business automation, was built by a separate team, minimizing cannibalization and allowing for rapid, focused development. It has since seen significant growth.
  • The Drag-and-Drop Dilemma: A more challenging pivot involved Retool’s core drag-and-drop application builder. Initially, the team doubled down on this approach, believing LLMs could be trained to use their existing tools. This led to the launch of a product in August last year allowing LLMs to use Retool’s interface.
  • The Courage to Course-Correct: David admits this was a mistake. Today, Retool is re-architecting its core product, focusing on having LLMs generate code that runs on the Retool platform, rather than having LLMs use their existing frameworks. This bold decision acknowledges the rapid evolution of AI and the need to adapt, even when it means rethinking years of development effort.

The Power of Clarity: Leadership and Decision-Making in a Fast-Changing World 🧭

David emphasizes the crucial role of clear, decisive leadership, especially in a rapidly evolving tech environment. He draws inspiration from Brian Chesky of Airbnb, who highlighted that decision-making is not zero-sum.

  • Empowerment Through Decision: By making clear, sometimes difficult, decisions, leaders empower their teams to move forward with confidence. Ambiguity, conversely, can paralyze an organization.
  • Simplicity Scales: A company’s strategy must be simple and communicable. If a strategy can’t be explained in a sentence or two, it’s likely a flawed one.
  • Embracing Imperfection: Making a decision quickly, even if it’s not perfect, is often more beneficial than prolonged indecision. The key is the self-awareness to course-correct when necessary.

The “Unsexy” Advantage: Focusing on What Customers Truly Need 💎

In a market flooded with tools promising productivity gains, Retool distinguishes itself by focusing on a unique value proposition that resonates deeply with its customers.

  • Not a System of Record: A deliberate choice was made not to be a system of record. This means Retool doesn’t store operational data. Instead, it empowers users to build applications on top of their existing data, wherever it resides (e.g., DataBricks, PostgreSQL, Salesforce).
  • Unlocking New Possibilities: This approach has been instrumental in attracting major clients like the US Air Force, Navy, and Army, who can maintain data security and control.
  • Beyond Developer Time Savings: The value proposition extends beyond simply saving developer hours. Retool enables the creation of hundreds, even thousands, of applications that would otherwise never be built due to prioritization or perceived complexity. One Retool application, for instance, reportedly saved a company $50 million by enabling the creation of a tool that wouldn’t have been prioritized otherwise.

The Rise of the AI Employee: Inevitable and Exciting 🤖

David firmly believes that AI employees are not a distant possibility but an inevitable future, driven by both technological advancements and economic pressures.

  • Economic Imperative: Businesses face immense pressure to optimize costs. With labor being a significant expense, cheaper AI alternatives will naturally emerge.
  • Technological Readiness: LLMs are already incredibly capable in knowledge-based tasks. The missing piece – “arms and legs” – is being addressed by products like Retool Agents, which enable LLMs to perform actions within a business.
  • Colgate’s Vision: Even traditional, large companies like Colgate are envisioning AI employees by 2028, signaling the widespread adoption of this trend.

Orchestrating the AI Workforce: Governance and Federated Models 🌐

As AI agents become more prevalent, the challenge shifts from writing software to managing it. David highlights the critical need for governance and a federated approach to AI adoption.

  • The “Last Mile” Challenge: The complexity lies in managing, deploying, and governing software, especially autonomous AI agents.
  • Federated Model for Success: The most successful approach combines centralized foundational elements (like data access and action layers) with decentralized empowerment for teams to build.
  • The Sprawl Problem: Without centralization, companies risk “sprawl” – multiple teams building similar, uncoordinated, and potentially incorrect solutions, as illustrated by a customer’s experience with multiple teams building identical Retool apps with different data.
  • Balancing Guardrails and Creativity: A federated model with clear guardrails allows for centralized control while fostering the creativity of individual teams who understand their domains best.

Navigating Enterprise Constraints and the “Boring” Path to Success 🛤️

David challenges the notion that enterprise constraints, like a reliance on Microsoft products, are insurmountable barriers. He advocates for a “boring” but highly effective strategy focused on delivering differentiated value.

  • Differentiated Value Proposition: Even in environments dominated by major tech players, a compelling and differentiated product can gain traction. Government entities like the US Army, Navy, and Air Force, despite being Microsoft shops, are Retool customers due to Retool’s unique value proposition.
  • The “Deploy Anywhere” Advantage: Retool’s ability to deploy in any cloud environment (AWS, Azure, GCP, on-prem) significantly eases adoption for large enterprises.
  • The Power of the “Boring”: Retool’s focus on internal tools, often perceived as “unsexy,” has proven to be a strategic advantage. More than half of the world’s software is internal-facing, and by serving this often-overlooked market, Retool has built a sustainable and dominant position. This contrasts with some AI companies experiencing rapid revenue growth but struggling with high churn rates due to a lack of sticky, fundamental use cases.

The Product Leader’s Imperative: Building for the Future Today 🔭

David underscores the vital role of continuous building and prototyping for product leaders, especially in a rapidly evolving industry.

  • Seeing the Future: Product leaders must anticipate technological shifts and market needs, not just for the next quarter, but for years ahead.
  • Hands-On Prototyping: Experimenting with new tools, like cloud code and AI agents, provides invaluable insights that directly inform product development. David spends significant time building, using these experiences to shape Retool’s agent product.
  • The Necessity of Reinvention: In a dynamic landscape, constant iteration and self-reinvention are not optional; they are essential for survival and success. As David aptly puts it, “If you’re not reinventing yourself, someone else is going to reinvent you.”

This conversation with David Hsu offers a powerful masterclass in strategic thinking, product development, and leadership in the age of AI. It’s a reminder that even in the face of rapid change, a clear vision, a commitment to building, and the courage to make bold decisions are the cornerstones of lasting success.

Appendix