Principles

The Data Guild is a venture studio based in San Francisco, California. We help to bring data-driven products to market - whether via spinouts from the Guild itself, by working in the trenches alongside portfolio company teams, or in concert with strategic partners. We work in healthcare, life sciences, renewable energy, and climate change – complex sectors whose needs are not well-served by the existing venture ecosystem, and in which finding innovative solutions and bringing them to market is crucial for the health and sustainability of our planet and society.


We are steeped in machine learning, data science, and artificial intelligence, but we are adamant that technology choices be driven by a deep understanding of the problem at hand. In order to deliver real-world impact, we rely on the following principles in our approach to new product and business development.

User-centric design

The user and their experience is frequently, and unfortunately, an afterthought in data-driven product development. We start with that desired experience, and bring appropriate technologies to bear in service to it.

Systems thinking

In the complex and challenging domains in which we operate, the problem space must often be enlarged to have any hope of achieving the desired result. We always take a step back to think about the diverse forces that may be at work to encourage or inhibit change.

Data lifecycle

All too often, the data that is made available to address a problem is viewed as fixed, and obtaining the desired result can feel like squeezing blood from a stone. Before invoking sophisticated machine learning, we always look for ways to collect additional data, or otherwise augment our knowledge.

Hybrid intelligence

Fully-automated systems may get the attention, but systems that combine human and machine intelligence in thoughtful ways are often what gets the job done. We look to leverage the strengths of both kinds of cognition to create systems with capabilities that exceed what either can accomplish alone.