Our Investment in Refuel
.svg)
At the end of the day, it's all about data. I've been in the industry a long time and it's remarkable how the heart of many conversations with senior leaders is around data: the need to leverage data, have the right data, have the data shaped in the right way, get the data into the right hands, ensure applications are more data-informed. These are the topics circulating. This has been true for years. But now people are asking - what is the role of data in this AI wave that has just begun? Will the rise of AI and LLMs somehow make a business’s data less valuable?
It’s just the opposite. We believe that the companies that are able to instrument their businesses and use that to feed the coming generation of AI systems are going to not just be the ones to thrive, but the only ones to survive.
But how is that going to work and scale? Anyone who has mucked around in the reality of enterprise data management knows how messy the state of business really is with legacy systems, newer software systems, digitally instrumented physical systems, etc. None of these systems produce perfect, clean data sets. So if the instrumentation of my own business’s foundation needs to advantageously feed the new AI systems, and the data is inherently disorderly, what’s to be done?
It’s perhaps somewhat Inception-like, but the answer is AI. The way through this is to task AI systems to annotate the data in order to create usable data sets for model training, fine-tuning, model refinement, and even just inference.
This is what the team at Refuel is working on – they are working toward building the product that will label enterprise’s unique data sets and turn them into fuel for the downstream AI workflows across building, managing, and using models.
When we first met Rishabh and Nihit it was clear they saw a bigger picture. They were not just relying on their extensive backgrounds as ML and AI practitioners, but they had performed thorough customer research on how a broader set of businesses are going to need to approach building with AI. This foresight and looking beyond the smaller set of digitally-native companies that had been on the leading edge of adoption, is foundational to how they are approaching building Refuel. Refuel goes beyond the models, notebooks, deployment tools, and governance and is seeking to connect the AI platforms to the raw material of the business – their data.
The team here at General Catalyst is incredibly excited to be working with Rishabh, Nihit, and the whole team at Refuel. As the founders so insightfully put it, we are entering an era of AI abundance – but how that is all stitched together is dependent on the specific business and how it runs. Fundamentally, the business’s data determines this – so getting scale and leverage from that data is of first-order importance. This is what Refuel’s first product is looking to do.
At the end of the day, it's all about data. I've been in the industry a long time and it's remarkable how the heart of many conversations with senior leaders is around data: the need to leverage data, have the right data, have the data shaped in the right way, get the data into the right hands, ensure applications are more data-informed. These are the topics circulating. This has been true for years. But now people are asking - what is the role of data in this AI wave that has just begun? Will the rise of AI and LLMs somehow make a business’s data less valuable?
It’s just the opposite. We believe that the companies that are able to instrument their businesses and use that to feed the coming generation of AI systems are going to not just be the ones to thrive, but the only ones to survive.
But how is that going to work and scale? Anyone who has mucked around in the reality of enterprise data management knows how messy the state of business really is with legacy systems, newer software systems, digitally instrumented physical systems, etc. None of these systems produce perfect, clean data sets. So if the instrumentation of my own business’s foundation needs to advantageously feed the new AI systems, and the data is inherently disorderly, what’s to be done?
It’s perhaps somewhat Inception-like, but the answer is AI. The way through this is to task AI systems to annotate the data in order to create usable data sets for model training, fine-tuning, model refinement, and even just inference.
This is what the team at Refuel is working on – they are working toward building the product that will label enterprise’s unique data sets and turn them into fuel for the downstream AI workflows across building, managing, and using models.
When we first met Rishabh and Nihit it was clear they saw a bigger picture. They were not just relying on their extensive backgrounds as ML and AI practitioners, but they had performed thorough customer research on how a broader set of businesses are going to need to approach building with AI. This foresight and looking beyond the smaller set of digitally-native companies that had been on the leading edge of adoption, is foundational to how they are approaching building Refuel. Refuel goes beyond the models, notebooks, deployment tools, and governance and is seeking to connect the AI platforms to the raw material of the business – their data.
The team here at General Catalyst is incredibly excited to be working with Rishabh, Nihit, and the whole team at Refuel. As the founders so insightfully put it, we are entering an era of AI abundance – but how that is all stitched together is dependent on the specific business and how it runs. Fundamentally, the business’s data determines this – so getting scale and leverage from that data is of first-order importance. This is what Refuel’s first product is looking to do.
.jpg)
