Blog

Numbers Station Interview with Mark Nelson, former CEO of Tableau

Mark Nelson, former CEO of Tableau, shares insights with Chris Aberger, CEO of Numbers Station, on the past, present, and future of data analytics. They discuss the evolution of the modern data stack and the gaps that remain, traditional and new roles in data and analytics, and the future of data analytics with foundation models.

Introducing Numbers Station Labs

Numbers Station Labs is a research lab dedicated to bringing foundation model technology to the modern data stack. The lab was founded by a group of PhDs from the Stanford AI lab where the term foundation models (FMs) was coined. Our mission is to remove the high barrier to entry currently associated with the modern data stack and to open up new possibilities for enterprise data automation. We believe that continued dedication to research-based innovation is key to their impact on the modern data stack and the success of their customers. Learn more about our research agenda.

Prototype then Deploy: Shrinking Foundation Model Quality into Small Models

Foundation Models (FMs) such as GPT-4 promise a natural language interface for data analysts to perform complicated transformation tasks and gain insights from their data. Despite the wide success of FMs, there are still high barriers to use on enterprise data such as hard to prompt, privately owned, and a high cost to run (a single run of 1M rows could cost more than $3.7K!). At Numbers Station, our goal is to bring state of the art FM technology to all data analysts. With a few natural language examples from an analyst, Numbers Station customizes a small FM on the analyst’s task and data. Our smaller FMs match the capabilities of proprietary large FMs and can be up to 2000x times cheaper. At Numbers Station, we are excited to take the first step towards making FM technology accessible to all enterprise analysts.

SQL Coding Assistants Customized to Enterprise Logs

Numbers Station wants to bring the productivity of foundation models (FM) for code to the data analyst and help them generate SQL to transform their data. SQL FMs need to be heavily personalized to the data and infrastructure the analyst interacts with, but customizing FMs is difficult because they are difficult to prompt, prohibitively expensive to finetune over enterprise data, and not applicable over private data. Numbers Station introduces Data Transformation Assistant to help the enterprise customize their own smaller SQL FM that can be deployed locally. Numbers Stations produces competitive models on two benchmark datasets – even outperforming GPT-4 - and can be more than 2,000 times cheaper than conventional approaches.

Numbers Station x dbt: built for analysts, on the tools engineers love

Available now, the Numbers Station Data Transformation Assistant democratizes the data transformation process to all skill levels with foundation model powered technology. This blog explains why we decided to invest in dbt, and how Numbers Station enables data analysts to create efficient and accurate analytics pipelines.

Ask Me Anything: A simple strategy for prompting foundation models

Foundation models, due to their size, can be prohibitively expensive to deploy and difficult to adapt to enterprise data tasks. This blog describes our recent research work, in collaboration with the Stanford AI lab, to deploy cost-effective, smaller foundation models that match or exceed the performance of 30x larger foundation models.

Can foundation models wrangle your data?

‍State-of-the-art data wrangling is not self-service, leading organizations to use legacy, slow-moving technology. Using cutting-edge foundation models, Numbers Station lowers the barrier to state-of-the-art data wrangling, accelerating existing wrangling processes and enabling data analysts to derive new insights from their toughest data.

Numbers Station: Intelligence for the modern data stack

Numbers Station is building Foundation Model-powered technology to bring automation in data-intensive enterprise workflows, automating tasks so data workers spend less time on mundane data tasks, and more time on the work that matters.