September 2025 Summaries

AI agents and AI workflows are both automation methods using generative AI, but they differ significantly in their structure and execution. AI workflows operate with a predefined set of steps that yield consistent and predictable results, ideal for tasks with well-defined inputs and outputs, such as support ticket routing or data processing. In contrast, AI agents function autonomously, deciding on actions and adapting to new information or conditions, making them suitable for open-ended or ambiguous tasks like research or customer support. While AI workflows are easier to implement and debug due to their deterministic nature, AI agents offer greater flexibility and can handle complex decision-making and dynamic scenarios. The choice between using AI agents or workflows depends on the specific needs of the task, with hybrid approaches often providing a balance of predictability and adaptability. As AI technology evolves, the trend is moving towards integrating both approaches to address diverse business challenges effectively.

Aaron Schnider, a finance analytics engineer at Lithic, exemplifies the modern approach to building production-ready applications by using AI coding tools and his unique expertise in finance and data. Transitioning from a CPA to a role at a digital asset exchange, Aaron leveraged his background to develop a custom billing application named Bill Buddy, which addresses inefficiencies in back-office processes by offering a user-friendly interface while ensuring data integrity. His work highlights the growing trend of domain experts creating tailored solutions without relying solely on engineering teams, thanks to AI tools that enable rapid prototyping. Aaron emphasizes the importance of stakeholder empathy, which allows him to bridge the gap between technical and business teams, ensuring that applications like Bill Buddy are both technically sound and aligned with business needs. His approach demonstrates how domain experts can harness their understanding of specific processes to build solutions that are reliable, secure, and capable of evolving with organizational needs, ultimately enhancing productivity and innovation within their fields.