Meet the Agents at USV: Arthur, Ellie, Sally, and Friends

In the world of technology, innovation often thrives in unexpected places. A prime example of this is the story of Building 20 at MIT, a temporary wooden structure built during World War II to house the urgent classified development of radar technology. As described in Stewart Brand’s "How Buildings Learn," this hastily constructed three-story building was meant to be a temporary solution, but it ended up becoming one of the most cherished and productive buildings on campus for 55 years. The flexibility and adaptability of the space allowed researchers to shape it according to their needs, knocking down walls and modifying the building to fit their workflow.

Fast forward to the present day, and the concept of adaptable, temporary structures has found a new life in the realm of software and artificial intelligence. Over the past year, AI agents have enabled teams and individuals to create customized tools that align with their unique ways of working. This Cambrian explosion of custom agents, tools, and software is transforming how different teams operate, from engineering teams building internal coding agents to non-engineering teams adopting similar practices.

At Union Square Ventures (USV), we have been at the forefront of this movement, creating our own internal agents and custom software to fit how we work. In this article, we will explore the learnings we have gained from our recent endeavors over the past three months.

Our journey began with a simple yet profound question: How can we solve one problem effectively? The answer, we discovered, lies in focusing on a single challenge and building a solution around it. By doing so, we can create a more targeted and efficient tool that addresses the specific needs of our team. This approach has been instrumental in the development of our AI agents, as it allows us to tailor them to the unique workflows and requirements of our diverse teams.

One of the key insights we have gleaned from this process is the importance of collaboration. To create effective agents, it is crucial to involve team members from various disciplines, such as engineering, design, and operations. By fostering open communication and collaboration, we can ensure that the tools we build are not only technically sound but also user-friendly and aligned with the needs of those who will be using them daily.

Another important lesson we have learned is the value of iterative development. In the fast-paced world of technology, it is essential to be agile and able to adapt quickly to changing needs and feedback. By breaking down our projects into smaller, manageable tasks and continuously refining our solutions based on user input, we have been able to create agents that evolve and improve over time.

In addition to these lessons, we have also discovered the power of leveraging existing tools and frameworks. By building upon well-established platforms, we can create custom agents that are both efficient and scalable. This approach not only saves time and resources but also ensures that our tools are compatible with the technologies and systems already in use by our teams.

As we continue to innovate and adapt, we are excited to see the potential of AI agents to transform not only our own workflows but also those of our partners and clients. By harnessing the power of customization and collaboration, we believe that we can create tools that not only enhance productivity but also foster a sense of ownership and engagement among team members.

In conclusion, the story of Building 20 at MIT serves as a powerful reminder of the value of adaptability and innovation. By applying these principles to the world of software and AI agents, we are creating tools that not only solve specific problems but also empower teams to work in ways that are unique to their needs and workflows. At USV, we are committed to continuing this journey, building custom agents and software that will shape the future of how we work together.