2021 in Database Startups: Gold Rush

Maybe not exactly the same bubble, but it's hard not to make parallels. Pets.com, one of the most comical cases of 2000s, was worth $400M at peak. When Chewy.com went public, they dwarfed that number with a new record of $13B. The analogy was so obvious, that CNBC wrote an article on that [1].

After that $CHWY managed to go as high as $118/share, translating into a market cap of roughly $50 Billion in December 2021. Now they are trading at about half that price.

[1]: https://www.cnbc.com/2019/07/26/opinion-chewy-is-no-petscom....

I'm not sure I agree with this - there have been many postgres features over the past ~10 years that got me excited, some of the highlights that immediately spring to mind:

- JSON / jsonb

- improved partitioning support

- identity columns

I agree that I wouldn't want to trust a new entrant to the space unless it was subject to rigorous testing, eg https://jepsen.io/

> You need to evangelize a product on which the life of someone’s project or company depends.

I've actually seen the results of a database product being picked, the company behind which ceased to exist. It wasn't pretty, since at one point the product ceased to be supported and therefore neither any updates/fixes were made, it wasn't available in the repositories for new OS distros and eventually even the documentation for it went offline. Having to support a system that integrated with it was an unpleasant experience, all the way to it being eventually replaced with something else instead.

Therefore, it probably makes a lot of sense to base something as critical as your data storage layer on proven technologies that have demonstrated that they'll probably be supported in one form or another for the following years or even decades, unless you have a good reason for choosing something else.

Those reasons might deal with particular workloads or requirements, e.g. clustering solutions for PostgreSQL/MySQL/MariaDB/..., geospatial extensions, solutions to integrate with it through REST interfaces or even GraphQL or something like that, with a stable and proven piece of software still at the core of it all.

In case anyone is wondering, the product in question was Clusterpoint, about which you can read a bit more here: https://en.wikipedia.org/wiki/Clusterpoint a NoSQL database that actually predates MongoDB by a few years, as far as i know. Of course, now it seems like even their homepage is offline.

> The biggest lie I have seen across databases is actual time travel for your data. Or in other words given a date and time in the past it gives you what the data was then.

I worked at a company where we built this for our internal HBase clusters. Given any timestamp (or even better a commit ID) we could restore the cluster to that point in time. This was done through a combination of backing up the store files + write ahead logs. As I recall we had a retention limit for how much time we could do this for - but they were high enough for all of the "oh crap" moments.

While not a database. Apache iceberg has the time travel functionality for read heavy big data analytics, among other optimizations.

On the business model side of things, I think the most important thing to think about is the threat of the cloud providers. In the same way that MSFT were able to take most of the profits from the 90s/00s PC business, and Apple were able to take most of the profits from the appstore, and Amazon takes most of the profits from the Amazon market place, in the same way the cloud providers are realistically going to take the vast majority of the profits from running your DB in the cloud. Elasticsearch has exactly this problem. If they're lucky, Amazon will buy them. If they're unlucky Amazon will just steal their tech and take the profits (some would argue this already happened). They're by and large living in someone elses ecosystem.

Thus the reason they have created new licenses that poison the well for companies trying to offer SaaS services using mongo or elastic as two examples or just GPL for Neo4j

Yes it is! It’s the first link in this article as well)

Thanks a lot! I have never worked with RethinkDB specifically, but I can assure you - it's only one of many names in the DBMS graveyard. Most have probably died because of inability to monetize open-source software.

For DBMS startups, which don't offer an order of magnitude improvements, the problem is convincing any one to replace a fundamental piece of infrastructure. For disruptive startups the problem is even bigger. As soon as you go open-source, all of your competitive technological advantage is gone.

> it seems they’re able to significantly beat the incumbents, partly by collapsing down the layers.

Hahaha, thanks for the kind words! Not sure if I could describe our methodology better)) We originally come from and hope to continue working on AGI technologies (Artificial General Intelligence).

Writing a key-value store and, subsequently, a DBMS came out of necessity, as other solutions seemed too bulky, slow, outdated and expensive. So we invested time into building UnumDB for internal use, but after first benchmarks, it was clear, that this must be available to broader audience. We will polish it in the next couple of months and make it deployable in public clouds, just like other DBMS brands.

Now DBMS is our central focus, but with essential R&D out of the way, we will rebalance our priorities and continue working on BLAS and sparse neural nets.

Seemed like they had some traction in MSFT shops 5-6 years ago, guessing they hit some synergies in the .NET ecosystem?

Thanks!

It was on the longer list of smaller DBMS brands, but I was afraid to make the volume incomprehensible. Maybe we will write a more focused version, covering just smaller startups in 2022.