Superblocks CEO: How to find a unicorn idea by studying AI system prompts

Superblocks CEO Brad Menezes suggests that analyzing system prompts from leading AI startups offers valuable insights into prompt engineering, highlighting role prompting, contextual prompting, and tool use as key areas of focus. By studying these prompts, Superblocks gained learnings that informed the development of their AI coding agent, Clark, and identified an opportunity to cater to non-programmers by addressing security and enterprise data access. Superblocks, which recently raised a $23 million Series A, leverages its vibe coding tools internally, enabling non-technical staff to build internal tools and agents, with customers including Instacart and Paypaya Global.

Analysis

Superblocks, an enterprise 'vibe coding' startup, has secured a $23 million Series A, elevating its total funding to $60 million, to develop tools enabling non-developers to build AI applications. CEO Brad Menezes highlights that while system prompts from existing AI unicorns offer valuable learning in prompt engineering—particularly concerning role definition, context provision, and tool utilization—they represent approximately 20% of an AI product's core value. The remaining 80%, termed 'prompt enrichment,' encompasses the critical infrastructure, data integration, and post-processing logic built around foundational model calls. This insight shapes Superblocks' strategy to differentiate by managing complexities like security and enterprise data access, including integrations with platforms such as Salesforce, thereby targeting the non-programmer segment. Early traction is evidenced by customer acquisitions like Instacart and Paypaya Global, and the company's internal use of its own AI agents for various business functions underscores the practical utility and potential for democratizing AI tool creation within enterprises. The viral interest in Superblocks' shared system prompts further indicates significant market attention on prompt engineering techniques and AI development methodologies.

Key Decisions for Investors

• Investors should evaluate the expanding market for AI-driven low-code/no-code platforms targeting non-technical enterprise users, as exemplified by Superblocks' recent $23 million Series A funding and enterprise client acquisitions like Instacart (CART), which signals a potential growth vector and competitive shift in enterprise software.
• When assessing AI startups, prioritize evaluation of their 'prompt enrichment' capabilities, proprietary infrastructure, and enterprise integration strategies, particularly with systems like Salesforce (CRM), over sole reliance on prompt engineering, as these elements constitute the primary drivers of defensibility and value.
• Monitor the trend of enterprises empowering business units to independently develop AI agents for internal tasks, as this indicates a shift towards decentralized IT development and could drive demand for specialized, user-friendly AI application builders, impacting operational efficiency and IT spending patterns.
• Consider the indirect positive implications for foundational AI model providers, including Alphabet's Google (GOOGL), as the proliferation and success of application-layer companies like Superblocks, which leverage these core technologies, can stimulate broader adoption and validate the utility of the underlying AI platforms.