October 2025 Summaries

Assist, a feature from Retool's AppGen capabilities, was developed with a focus on minimizing friction and ensuring precision in building internal tools that businesses depend on, rather than merely optimizing for speed. Initial user concerns about data sourcing and potential costs highlighted the need for a system that engages users in decision-making, ensuring the correct organizational context is captured through human guidance. Four principles were established: building with users rather than for them, optimizing code for human readability, leveraging enterprise foundations for security and compliance, and maintaining transparency about the tool's limitations. Assist's design allows for a collaborative approach where AI assists but does not replace human input, ensuring applications are both reliable and maintainable. Since its launch, Assist has seen rapid adoption, with a significant portion of new apps being built using its capabilities, confirming the effectiveness of these principles.

In a fireside chat at the Retool Summit, Patrick Collison and David Hsu discussed various topics, highlighting the importance of pragmatism in software development and the broader implications of human progress. Collison shared insights from his early programming experiences, emphasizing the need to prioritize practical solutions over elegant ones, which influenced Stripe's foundational philosophy. The conversation delved into why companies persist with inefficient tools like spreadsheets, despite their limitations, and how Stripe overcame these challenges through investments in internal tooling and automation, such as their AI-driven "minions." The discussion also touched on the potential of AI not as a replacement for human workers but as a means to expand their capabilities, drawing parallels with the benefits consumers and society gain from advancements in technology. A key theme was the shift towards making software more flexible and customizable, echoing early computing visions of empowering users with greater agency. The session concluded with a focus on balancing software malleability with maintainability, advocating for a development approach that encourages innovation through well-defined constraints, as exemplified by Retool's AppGen capabilities.

Retool's first Summit, held on October 7 at SFJAZZ, showcased its new enterprise AppGen platform, designed to bridge the gap between prototype and production in software development. The event highlighted Retool's capabilities in enabling faster app development by integrating AI and natural language to generate production-ready applications that connect directly with real data, maintain security, and fit within existing systems. Key themes included the challenges of scaling AI safely, democratizing app creation across organizations, and defining the roles of humans and AI in software development. Presentations from companies like Uber and Komatsu illustrated the platform's impact in solving specific operational challenges with small, high-impact applications. The summit concluded with discussions on the evolving role of developers as system architects and the importance of governance in enabling widespread, safe innovation. The AppGen platform was praised for empowering non-technical users to build meaningful applications, marking a significant shift in enterprise software development by aligning the speed of innovation with robust governance and security.

Retool has introduced an AI-assisted app-building experience that aims to streamline the creation of production-grade software by integrating business data while ensuring data security and governance. This new feature allows users to leverage AI within the Retool editor by starting with pre-written prompts or creating their own, providing a balance between AI-driven development and manual control over application components. The Assist panel in Retool is a versatile tool that facilitates app generation, modification, debugging, styling, and understanding, by allowing users to use natural language prompts to handle various app-building tasks. Additionally, Retool ensures security by automatically inheriting resource and query permissions, making it possible to develop apps without extensive security reviews for each instance. The platform supports a hybrid approach by combining AI-generated scaffolding with manual coding for customization, allowing developers to efficiently build, modify, and understand apps while maintaining control over specific elements. Retool's new functionality is accessible to both cloud and self-hosted customers, promoting a new era of software development by making app generation more accessible and efficient.

Retool's enterprise AppGen platform is revolutionizing software development by empowering both domain experts and engineers to build customized solutions using AI, while adhering to enterprise standards. The platform's AI capabilities are designed to understand existing applications, allowing for collaborative development where AI assists users without replacing them, thereby ensuring production-quality outputs. It utilizes a format called ToolScript to make applications LLM-readable, facilitating intelligent edits by managing context efficiently and ensuring data connections are handled securely. By implementing a human-in-the-loop approach, Retool maintains control over changes and enhances learning opportunities, all while prioritizing security with features like SSO, RBAC, and data-level permissions. This ensures that applications are generated with a structure that respects users' permissions and data security, allowing for safe and efficient AI-assisted app building on users' existing cloud infrastructure.

Retool has launched the world's first enterprise application generation platform, AppGen, which enables users to create applications using natural language directly on their data, such as databases and APIs, with built-in security and governance. This platform resolves the longstanding trade-off between the speed of packaged tools and the flexibility of custom code by leveraging generative AI to create relevant, production-ready software that understands enterprise data and security contexts. Retool's AppGen, including its new feature Assist, allows both domain experts and engineers to collaborate on the same platform, enhancing the speed and autonomy of software development while maintaining enterprise-level security and compliance. This democratization of app building empowers non-engineers to solve problems and implement solutions rapidly, transforming how organizations operate by increasing the number of builders and accelerating the delivery of solutions. The platform is designed for scalability, composability, and security, promising to reduce the time needed to move from questions to actionable software solutions.

Software development is undergoing a transformative shift as non-engineers increasingly build production-ready applications using enterprise AppGen tools, which combine low-code capabilities with AI. This democratization of software creation enables operations managers, product leaders, and finance analysts to independently develop solutions that meet their teams' needs, significantly reducing development cycles from months to days. The survey of 1,128 Retool builders reveals that 65% of participants are non-engineers who now contribute to software development, highlighting a blurring of traditional roles between technical and business functions. This shift is driven by AI productivity mandates, with 66% of builders working under such directives, leading to a new economic model where custom solutions are created more efficiently and cost-effectively. Despite initial concerns over AI mandates, the benefits include increased productivity and the ability for organizations to scale horizontally without relying solely on hiring or external software vendors. However, the success of this model depends on implementing guardrails within platforms to ensure security and compliance, transforming business users into pivotal problem solvers and driving organizational growth.

Vibe coding, a method of creating AI agents using natural language instead of traditional programming, is explored through the platform Retool, which allows developers to build agents by describing their functions in plain language. Retool's Configuration Assistant helps transform these natural language descriptions into fully configured, production-ready agents by automatically setting up the necessary tools and instructions. The platform provides a structured approach to agent development, incorporating observability, governance, and workflow management to ensure reliability and scalability. While vibe coding offers flexibility and rapid prototyping, it also presents challenges such as ambiguity and performance bottlenecks. Retool addresses these by integrating features like AI Action blocks for deterministic tasks and Invoke Agent blocks for autonomous decision-making within workflows. This approach allows users to efficiently build, test, and deploy AI agents that can handle complex tasks, such as managing meetings on Google Calendar, while maintaining control and reliability through built-in monitoring and governance tools.