Introducing Bobsled: The First Data Sharing Platform
Data cannot stop at the edge of a business. Critical insights about a company’s market are curated by providers, generated by SaaS apps and shared by customers and partners. Without access to this external data, data teams only see a part of the world in which they operate and will struggle to help inform better decisions.
But today, external data is still effectively locked up in silos. Sharing data between companies is an error prone, antiquated process that requires toil, creativity and heroics from both provider and consumer.
That’s because the infrastructure for cross-organizational data collaboration is anything but modern. Data providers frequently rely on tools either built in another decade (e.g. sharing files through SFTP) or designed for another purpose (e.g. using an API built for transactional sharing). As a result, data consumers are left to build and manage their own set of pipelines to pull the data back into their native environment – all before they can even begin to think about making it analytics-ready.
The result is a high-pressure, deeply fragile game of telephone that leaves no one happy and everyone wanting more.
The industry has not been blind to this challenge. In recent years, major data platforms like Snowflake and Databricks have introduced new capabilities (i.e. Snowflake Sharing and Delta Sharing) that dramatically improve the data collaboration experience. But importantly (and intentionally), these technologies only work when teams are within the same ecosystem.
The result is that leveraging external data often remains too expensive and too time-intensive to justify unless a company has a massive data team or use case with an unimpeachable path to value. At Bobsled, we think that should change.
The solution? A universal data collaboration layer
What data teams need is infrastructure purpose-built for external data collaboration. They need infrastructure to work on any platform, connect to any team, and support any type of data. They need infrastructure built with the entire workflow in mind: sharing and consuming data should be equally frictionless.
Bobsled was founded on these convictions. Looking at the rest of the ecosystem, we saw an urgent need for a truly independent, cross-cloud solution to drive collaboration that met five key design principles.
- Cross-cloud, not multi-cloud. There’s no one right data platform. Forcing companies to either limit sharing to others in their own platform is not practical; nor is asking them to have a presence in every one either.
- No more intermediaries, no more marketplaces. Stop building different places or ways to share data, and start making the existing ones more accessible. We do not need new sharing protocols or new marketplaces, we need better ways to collaborate within the spaces that already exist.
- A new category, not another feature. Collaboration can’t just be a feature of another data platform. It requires the full focus of a truly independent company: one without competing products or initiatives that might distract its focus or compromise its intentions.
- Software, not services. This problem fundamentally requires infrastructure. Services do not scale, do not generalize, and do not allow participants to collaborate in real-time.
- Collaboration cuts across the enterprise. Collaboration needs to work across personas. This means a complete solution is both embeddable as well as UI-based.
Introducing Bobsled – the Data Sharing Platform
After over a year in stealth, we’re excited to publicly launch Bobsled, the world’s first Data Sharing Platform. With Bobsled, data providers, SaaS companies and data teams can share data into another company’s cloud data lake or warehouse without ever leaving their own.
Bobsled is not a new sharing protocol. It’s a platform that allows data providers to capitalize on existing technologies from a single control plane. With Bobsled, providers simply connect to their source, pick the consumer bucket or warehouse where they want the data to go – and Bobsled handles the rest. Providers can deliver anywhere with zero additional development work and consumers get analytics-ready data natively in the data lake or warehouse where they work.
The response in the market, even in stealth, has exceeded our expectations. Companies like LinkUp, TrueData, Facteus and more are already using Bobsled to run their data delivery – and the results are transformational. What used to take days (if not weeks) now takes minutes and companies can move valuable engineering resources from managing infrastructure to building products.
"It used to take us 1-2 days to get our customers up-and-running on our datasets," said Toby Dayton, CEO of LinkUp, a leading provider of job market data. "With Bobsled, we can get a customer working on an analytics-ready dataset in the platform of their choice within 1-2 hours.
New Funding to Build a Faster Track and a Bigger Sled
Solving a big problem requires substantial investment and great partners. That’s why we’re also excited to announce that Bobsled has raised $24 million in Seed and Series A funding to make this vision a reality. The Series A funding was led by Madrona and Greycroft Partners and the previously unannounced Seed round was led by .406 Ventures.
The quick succession of funding rounds, even in a difficult VC climate, underscores the size and importance of the opportunity ahead of us. Our venture partners believe as we do that the data industry is on the verge of a monumental transition. We believe that data is one of the most valuable resources in the modern era and making it frictionless to access – no matter whether it’s from a different team or a different company – can be transformational.
Our belief is that in doing this we will help unlock a new generation of data companies. The groundswell of interest in data sharing, data-as-a-service, and data products all point to a future where companies' data becomes not just a source of insight, but a product itself. We’re committed to building the infrastructure to make this happen and cannot wait to see what others build.