As some of you know, I started building a small fund as a platform for the startup investing I enjoy doing. It is for now a small, manageable fund, and something I plan to grow and develop over the years. I am excited to be working with talented founders whose foresights will change the way we all work, and along the way generate great returns for early-stage investors.
I recently invested in two data companies. Both are exciting startups with potential to transform huge industries, led by strong teams who understand their markets. I am fortunate to be involved with such strong founders, investors, and partners.
Mode Analytics is an exciting data science platform. We are participating in a round, soon to be announced, with fantastic co-investing partners. The founding team Derek, Josh, and Benn are early members of Yammer's Analytics team, one of the strongest data science teams in the Valley. They have deep analytic and tech experience.
Mode makes data analysis easy by storing data scientists' work and resurfacing it when relevant. Imagine you are an analyst trying to understand average churn rates among your customers. Mode makes it easy to find any previous work done by other analysts related to churn and adapt it to your specific use case.
Data rework and collaboration are a huge pain points for most companies. And these problems are even greater in the open data community. Open data repositories don't have an incentive to contribute back and further the community. Mode offers a public product that is as an analyses repository (not a data repository). Mode plans to mimic GitHub's very successful spread from open-source into corporate mainstay.
PipelineDB is building a very powerful database technology. We are participating in around with great co-investors that include SV Angels, Paul Buchheit, and Data Collective. The team is led by two former Adroll employees: their head of data engineering, Derek, and their top enterprise sales rep, Jeff. They are graduates of the most recent YCombinator class.
Today’s technologies for realtime processing of large volumes of data in are complex and inaccessible to most engineers. PipelineDB offers a simple, scalable, realtime database that continuously processes incoming data. This eliminates the need to store excessive amounts of granular data, achieving extremely high levels of throughput on modest hardware.
PipelineDB does this with a simple query language and architecture. No application code is required. Fundamentally, PipelineDB is a unified solution to the two biggest problems in data today: (1) an increasing volume of data produced, and (2) an increasing need to analyze it quickly.
Derek and Jeff needed this product when they built it at AdRoll, and they have validated the demand for it with an impressive number of signed letters of intent to use PipelineDB.