Superbase's Open-Source Journey to Decacorn Status and AI Growth

Paul Copplestone, co-founder and CEO of Superbase, discussed the company's journey from an open-source tool for hosting PostgreSQL to becoming a decacorn and a significant player in the developer tools space. He shared insights into Superbase's origins, its open-source philosophy, developer experience, and adaptation to the AI era.

Early Entrepreneurial Journey and Superbase's Inception

Copplestone's entrepreneurial spirit is inherited from his father. After a brief and disliked stint in corporate consultancy, he committed to startups. Superbase is his third venture-backed startup, though he has been involved in several others. He clarified that startups often fail not primarily due to running out of money, but because founders give up or fail to find product-market fit.

His first startup, where he was CTO, struggled with product-market fit, leading to his departure due to differing visions with the CEO. The second startup was a temporary arrangement where he helped a co-founder launch a business while incubating Superbase as a side project. This company is still profitable, but Superbase became his long-term focus.

Superbase officially launched in 2020, but its roots trace back to Copplestone's experience with PostgreSQL and Firebase. After encountering scaling limits with Firebase in a previous company, he migrated to PostgreSQL and developed an open-source tool. This tool gained traction on Hacker News, providing the impetus to build a database startup.

The Choice of PostgreSQL and Open Source

Copplestone identified several reasons for choosing PostgreSQL:

• Open Source: He was committed to building an open-source company.
• Trending Technology: Monitoring Hacker News revealed PostgreSQL's growing popularity among developers.
• Reputation: PostgreSQL had a strong reputation, crucial for a database company.
• Game Theory Advantage: As an open-source project not owned by a single entity, everyone contributes to PostgreSQL, forcing hyperscalers to offer the best PostgreSQL services. This creates a "flywheel effect" of continuous improvement.

Initially, Superbase positioned itself as an "open-source Firebase alternative," a tagline that significantly boosted its adoption after an earlier "real-time PostgreSQL" positioning failed to gain traction.

The decision to be open source was philosophical for Copplestone and his co-founder, driven by a love for the community and building. Superbase is 100% open source, offering everything for free except the complex, proprietary code for hosting millions of databases, which is intertwined with billing logic and security.

Regarding the risk of large cloud providers hosting and reselling Superbase, Copplestone noted that they already host PostgreSQL databases. Superbase, however, is a stack of tools, making it difficult for others to replicate and manage effectively.

Monetization and Growth

Superbase was conceived with monetization in mind, though it took about 18 months to earn its first dollar. The team recognized that databases, while difficult to enter, offer significant growth potential once established. They targeted high-growth, venture-backed startups, particularly Y Combinator companies, as their ideal customer profile. Over 60% of every YC batch now uses Superbase.

Superbase's revenue growth was initially slow, reaching around $1 million ARR in its first year, then accelerating to $7.5 million the following year.

Developer Experience and Community

Superbase prioritizes developer experience by:

• Listening to the Community: Actively polling and engaging with developers on platforms like Reddit, Twitter, and Hacker News to understand their needs.
• Focusing on Time to Value: Early on, they measured the time it took for a user to launch a database and insert a row. Their goal was to reduce this from 8.5 minutes (for AWS RDS) to under a minute, eventually achieving 5 seconds. This focus on minimal input and rapid results is central to their developer experience.

The AI Shift and "Superbase for Platforms"

The advent of AI significantly impacted Superbase. In late 2022, they noticed a surge in database launches driven by AI agents from companies like Bolt and Lovable. Today, an estimated 60% (and likely closer to 90%) of their databases are launched by agents, with millions of launches monthly.

Initially, there was internal debate about supporting these AI use cases, as the early AI-generated apps were not always high quality, and their growth potential was uncertain. However, it became clear that their product was well-suited for platforms needing to launch millions of databases. This led to the creation of "Superbase for Platforms," a product that allows companies to manage and launch numerous databases. This product also appealed to enterprise innovation labs, which needed similar capabilities for managing internal databases, including security, monitoring, and lifecycle management.

The "AI shift" led to unprecedented growth for Superbase. While they don't spend on marketing, relying on word-of-mouth, their user base and conversion rates accelerated dramatically. The number of developers grew from 6.5 million to 10 million in a short period, with all key metrics trending upwards. This was attributed to AI agents enabling developers to build more and bigger applications faster and more successfully.

Evolving Developer Experience for AI Agents

The rise of AI agents as primary users has changed Superbase's approach to developer experience:

• Dashboard Evolution: Initially, the dashboard was interactive. With LLMs, it became more code-centric (SQL editor, chat interface).
• Agent-Centric Design: Agents prioritize CLIs and APIs. Superbase aims for a unified experience where the dashboard's functionality is mirrored in code.
• Infrastructure as Code: The goal is to represent the entire database in declarative schemas within code (e.g., a Superbase folder), enabling Git-based branching and multiple environments. The entire Superbase stack needs to be branchable to support agents in spinning up and managing complex databases from user intent.

Advice for Future Dev Tool Founders

Copplestone advises new dev tool founders to target agents from the outset. While it's challenging to become a default choice for agents (which might only select a few winners), founders should focus on:

• Ease of Use for Agents: Making it extremely easy for agents to build with their tools.
• Word-of-Mouth: Building a product that naturally generates positive buzz.
• Leveraging Open Source: Open-source nature helped Superbase by making its code and community discussions (on Reddit, GitHub) available for LLMs to "slurp up," effectively making Superbase known and used by agents even if human developers aren't directly aware.

He acknowledges the concern that this could lead to a few dominant players (like AWS), but believes that "vibes" and developer preference will still play a role, allowing companies like Superbase to compete.

Fundraising and Staying Grounded

Superbase has raised significant funding, including a $500 million round at a $10 billion valuation. Copplestone emphasizes that money should not change how a company operates; it's a tool to achieve existing goals.

To stay grounded, he believes founders must understand that money alone won't solve problems or guarantee success. This perspective allows smaller startups to compete with well-funded giants like AWS by focusing on other vectors of innovation and value.

His relationship with investors has been unique. Their initial seed investor was a hedge fund (Coatue) that became more interested as Superbase's success grew and public market potential emerged, a different dynamic from typical seed investors who might become less engaged after a return.

Distributed Team and AI Integration

Superbase operates as a fully remote company with 360 employees across over 60 countries, with no physical offices. Copplestone advocates for this model, emphasizing that it requires full commitment, not a hybrid approach.

A key advantage of their remote, "written-first" culture is that all decisions and communications are documented (on Slack, Notion). This wealth of written data is now invaluable for training internal AI agents, allowing them to access information and context from the company's entire six-year history, unlike most companies whose agents only have access to recent data.

Future Bets: Self-Driving Databases

Looking ahead, Copplestone believes the next big bet for Superbase is "self-driving databases." While current AI focuses on the "build stage" (prompting apps), the harder problem is the "operate stage"—how to continuously manage and maintain an application once built. This includes ensuring database uptime, security, patching, and overall operational resilience without human intervention. He sees this as a challenging but crucial area for innovation, especially as many current AI-driven startups focus on easily replicable "build stage" solutions.