Meet Bluesky CEO Mingsheng Hong, who is powering data cloud efficiency and innovation with Snowflake
Welcome to Snowflake’s Startup Spotlight, where we learn about innovative companies building their businesses on Snowflake. In this Q&A series, we ask startup leaders about the problems they are solving, how they use data, and how they are pushing the envelope in their respective industries. Stay tuned for some fascinating and inspiring conversations with the best and brightest startups!
California-based Bluesky is pioneering new ways of managing data by making querying and analytics faster, easier to adjust, and less expensive. Bluesky’s SaaS product uses an innovative technology called query patterns to analyze your workloads, then provides actionable recommendations to immediately optimize workloads and predictably stay optimized over time. This gives both data engineers and business leaders unprecedented visibility into data usage so they can focus on deriving business value from data rather than managing it.
We spoke to Bluesky co-founder Mingsheng Hong about the startup’s inspiration, the reasons behind its focus on optimization, and why the company is building on Snowflake.
Tell us a little bit about yourself and what inspires you.
I’m Mingsheng Hong, CEO and co-founder (along with Zheng Shao) of Bluesky. My career has been all about building best-in-class data and machine learning infrastructure to support some of the world’s most challenging workloads. I have more than 15 years of experience, most recently at Google, where I led teams to build storage and querying stacks powering Google’s $100 billion ads business and a new TensorFlow backend to improve performance for AI workloads.
As far as inspiration, I’ve been a data geek for about 25 years, way before the advent of big data. I’m inspired by the opportunity to truly democratize the power of data, analytics, and machine learning for any organization that wants to improve workload cost efficiency of cloud-based data platforms.
What problem does Bluesky solve?
Bluesky is designed to make querying and analytics faster and cheaper in order to deliver exceptional operational and financial value. Our approach to workload optimization enables data teams to leverage an intelligent and automated solution to quickly find optimal data layouts and warehouse settings.
Snowflake is famous for pioneering the Data Cloud and for championing consumption-based pricing. It’s innovative, and the architecture is incredible. It has delivered significant value in large part because of its SQL interface and how easy it is to get a project up and running. As companies increasingly run workloads in the cloud, it has become more challenging to properly attribute, monitor, and find surplus workloads to turn off to be more efficient. That’s where Bluesky comes in. We exist to make data efficient and simple, and to make data drive more business value, faster.
Why is solving this problem important right now?
Snowflake is such a powerful platform. Users no longer have to spend time provisioning heavy computer hardware upfront, which gives them unprecedented flexibility. Many users have experienced huge business and data growth over the past several years. But now we’re in an economic downturn, which naturally causes every CFO to review budgets and implement stricter guardrails. Cost governance, workload optimization, operational discipline, and cost reduction are becoming key priorities for Data Cloud users and finance executives. With a better understanding of costs and usage, a Snowflake customer can more effectively enforce accountability and improve the efficiency of its Data Cloud investment, while also ensuring that it has the cloud resources it needs to stay competitive.
Bluesky is purpose-built to holistically analyze Snowflake workloads and optimize them by proactively identifying potential issues like unoptimized queries, repeatedly failing queries, orphan tables, and so on. With Bluesky, you can eliminate data infrastructure inefficiencies—not only improving your Snowflake user experience, but freeing up engineering and financial resources that, in turn, can be funneled back into your business.
How is Bluesky doing things differently?
We go beyond basic infrastructure cost measurement. Bluesky looks at patterns in how customers use data across their entire instance, identifies opportunities for improvement, then provides granular insights by warehouse, cluster, users’ individual data pipelines, and queries to find opportunities to improve. We aim to eliminate work that human operators find tedious so they can focus on what they really care about: leveraging their data to drive immediate business value. We also know innovation is world-changing, so we help our clients get rid of wasteful spending so they can spend those resources on experimentation.
You’ve shared a lot about how Bluesky aims to help customers optimize their Snowflake instances, but you’re also using Snowflake at Bluesky. How has Snowflake enabled you to push the envelope in your industry?
Snowflake plays a critical role in the continuous operations of our SaaS product. We store and analyze our customers’ workloads through our Snowflake instance. We regularly use and benefit from Snowflake’s rich SQL support, extensive integration with other ETL/BI tools, and elastic cloud compute in both the product development and operations phases. Day-to-day data management with Snowflake has enabled us to focus our in-house resources more on our core business—finding and landing workload optimization ideas for our users.
Before Snowflake, we needed to invest more into setting up and maintaining cloud-based data infrastructure, which distracted from the investment into our core business. With Snowflake, we are able to focus our effort into developing our core product and growing our business.
What tech trend in your industry are you most excited about?
While the Data Cloud has been revolutionary, the majority of the world’s analytical data is still on premises. I’m most excited to work out how best to make efficient and sustainable use of data through a new generation of tools that help maximize data efficiency.
Additionally, because the industry is transitioning to usage-based pricing, I am excited about the emergence of “FinOps” as a new practice within an organization. The flexible concept of capacity that the cloud engenders is leading to a tactical shift and giving rise to a relatively new function for cloud cost management—FinOps—a framework to help companies monitor, manage, and optimize their cloud-based operating expenses. Bluesky is part of this trend and uniquely positioned to empower the modern, cloud-based enterprise to achieve its spend goals while furthering innovation.
What’s next?
Bluesky will focus on providing more intelligent, data-focused support to enterprises that have “cost with value” in mind. We have a great system for customers who need to observe and optimize the way they use data, and the next step in that journey is to fully operationalize things with active, intelligent monitoring; real-time alerts; decision intelligence; and automation.
We want customers to be able to automate safeguards that allow experimentation and innovation but prevent known wasteful spending. Why make boring and costly mistakes if you can avoid them? We want customers to feel confident they can experiment safely with data.
Most of all, we want to support customers as they move through every phase of Data Cloud use, from inception all the way to fully automated data cloud ecosystems. We aim to help them get the most out of data, whatever their goals. We hope one day we’ll all look back and wonder why we ever did things differently.
Learn more about Bluesky here, and check out the Snowflake Startup Program today.
Originally posted on October 26, 2022 @ 5:45 pm
