select star raises seed to automatically document datasets for data scientists
The Evolution of Data Access and the Rise of Metadata Management
In the early days of web development, a common practice involved retrieving all data from databases using “SELECT *” queries. This approach, while simple, is unsustainable in today’s data-rich environments.
Modern organizations grapple with petabytes of information, encompassing tens of thousands of tables and handling millions – even billions – of database requests. Consequently, data science teams can no longer efficiently request and process entire datasets immediately.
The Need for Data Warehouses and Discovery
The challenges of big data have spurred the development of data warehouses and data lakes – and more recently, data lake houses – designed to improve data accessibility and robustness.
However, a significant hurdle remains: data cataloging and discovery. Simply consolidating data in one location doesn’t guarantee that data scientists understand its meaning, ownership, or potential impact on existing web and corporate reporting applications.
Further reading on the evolution of data architectures can be found here: https://techcrunch.com/2021/02/04/a-lake-house-architect-a-miami-vc-and-homeowner-walk-into-a-wine-bar/
Introducing Select Star: Automated Metadata Solutions
Select Star addresses this challenge by automatically constructing metadata within a data warehouse environment.
The platform provides a full-text search capability, enabling users to quickly locate relevant data. It also utilizes “heat map” signals within search results to highlight frequently used columns and those with the highest number of referencing queries.
This product is delivered as a Software-as-a-Service (SaaS) solution, facilitating rapid onboarding by connecting to a customer’s existing data warehouse or Business Intelligence (BI) tool.
The Founder’s Vision and Experience
Shinji Kim, founder and CEO, developed Select Star based on her direct experience with data science teams in corporate settings.
Prior to founding Select Star (established in March 2020), she founded Concord Systems, a real-time data processing company acquired by Akamai in 2016. She observed that while data and computational power are readily available, understanding and utilizing the data effectively presents a new obstacle.
Kim emphasizes the increasing cost of “tribal knowledge” within growing companies and notes that larger organizations like Facebook, Airbnb, Uber, Lyft, Spotify, and others have already invested in building their own internal data discovery tools.
Her goal for Select Star is to provide a user-friendly platform that enables any corporation to efficiently solve this data discovery problem.
Funding and Competitive Landscape
The company recently secured $2.5 million in seed funding, led by Bowery Capital, with participation from Background Capital and several angel investors including Spencer Kimball, Scott Belsky, Nick Caldwell, Michael Li, Ryan Denehy and TLC Collective.
While data discovery tools have existed for years – with companies like Alation raising substantial venture capital – Kim believes Select Star can differentiate itself through a superior onboarding experience and increased automation.
Many existing tools require significant manual effort; Select Star aims to automate the process of establishing relationships between data elements, allowing teams to derive value more quickly.
Impact on Data Engineering and Database Management
Beyond simply understanding data, Select Star assists data engineers in managing database changes with greater confidence.
The platform can identify how columns are utilized and assess the potential impact of modifications on applications and other datasets, minimizing the risk of cascading errors.
Current Status and Future Development
Select Star has recently exited private beta. The company currently comprises a team of seven individuals and is focused on expanding its team and streamlining the user onboarding process in the coming months.