Company Updates

Pilosa + Molecula = FeatureBase: A Story of Evolution

A Bit of Background

The year was 2017. The #1 song in the US was Ed Sheeran’s “Shape of You.” The highest-grossing movie was Star Wars Episode VIII: The Last Jedi. Fidget spinners were making their first appearance. 

And in Austin, TX, a small, brilliant, and tenacious group of engineers was discovering a new way to store data using only bitmaps, and open sourcing their discovery.

Enter Pilosa, Stage Left.

For the next two years, this small but mighty team continued to build on their original discovery that the database didn’t actually need all of the “data” if you just stored relevant relationships between a record and a feature and represented that relationship with a bit. Over the course of these two years, the team was awarded 9 patents for their work on bitmaps. They grew the Pilosa community to include users around the world who also saw the performance power and I/O efficiencies that bitmaps can offer.

Molecula Appears, Downstage Right.

In 2019 (the #1 song in the US was “Old Town Road” by Lil Nas X, the highest-grossing movie was Avengers: Endgame, TikTok was just emerging as a trend, and Covid was but a fever-dream), the Pilosa team embarked on the creation of a commercial brand, Molecula, to bring its database built on bitmaps to market. The team continued to grow the product, achieving never-before-seen capabilities with bitmaps. This enterprise-grade version of Pilosa was named FeatureBase: a name derived from the way that the product represents data – broken down by individual feature (or attribute) with simply a 1 or 0 representing the presence or lack of presence of that feature within a record.

Following the advancements to our database, in 2021, the team made it a goal to create a cloud version of FeatureBase. The team succeeded, and FeatureBase Cloud was born as a way to get our extremely efficient data storage and computation format into the hands of users who do not want to manage all of the complexities of infrastructure/hardware. 

Every step we’ve taken to-date has been part of our journey towards building a database on bitmaps and sharing the performance and efficiency gains that it brings with the data infrastructure world. Along the way, we’ve focused on individual parts of our journey (community → enterprise → cloud) – that ends today.

Pilosa + Molecula = FeatureBase

The year is 2022 – the #1 song in the US is “Easy On Me” by Adele, the highest-grossing movie of the year is Top Gun: Maverick, and Covid BA.5 is the latest variant in our ongoing Covid saga.

And today, September 7, 2022,  we are happy to share that Pilosa and Molecula will be merging together into a single company called FeatureBase. To make it extremely clear:

  • Company: FeatureBase
  • Open Source Project: FeatureBase
  • Commercial Product: FeatureBase Cloud

Has this story been perhaps a bit long-winded and confusing because of all the different names that popped up? Likely yes. We feel it, too. That’s why we’re doing what we’re doing. Throughout all of our previous names, it’s always been the same team – a team of engineers that wants to create a community around the power of using bitmaps as the underlying storage format for a database. A team that believes in the efficiency of representing data as simply the presence of a relationship between a feature and a record. And that same team is ridiculously excited to dive back into open source, bringing some of our most up-to-date tech to the developer community, while continuing to build and maintain our Cloud offering for anyone who doesn’t want to tinker and tune.

Does it seem like we’ve zigged and zagged a bit? On a strictly cosmetic level, yes, but at our core, we’ve always been about creating the fastest, simplest, and most efficient analytical technologies possible. When we look back at the original Pilosa decks we used during GopherCon talks, Austin data talks, and even investor pitches, we’ve always talked about creating a distributed database built on bitmaps that allows for “accelerated queries across multiple, massive datasets.” And that’s still the bread and butter of what we do and what we’re building for.

There’s a saying that it takes 10 years to build a database. We’re on year 5, and we’ve already made monumental progress towards creating new and interesting ways to store and compute data with far more efficiency than any other database in the market today. For the next 5 years, we’re going to continue building on that foundation, supporting the community that has become just as fascinated by the power of bitmaps as we are, growing that community so that others can experience the benefits that bitmaps bring to analytics, and creating the fastest, simplest, and most efficient analytical technologies possible. We hope you’ll join us on our journey.