Natalie
Vais

Natalie Vais has been inside the machine. Not metaphorically - literally. She wrote production code on databases at Oracle, shipped Google Cloud Firestore as a product lead, and directed product at an analytics startup that Twitter eventually bought. By the time she joined Spark Capital as a General Partner in 2023, she had the kind of ground-level fluency that most VCs acquire through asking questions at conference panels. She got it by building the actual systems.

That track record shapes everything about how she invests. Her portfolio at Spark - TigerBeetle, Polar Signals, MotherDuck, ElectricSQL, PostgresML, Eppo - reads like a reading list from a very serious database engineer's bookshelf. These are not consumer apps. They are the invisible infrastructure underneath apps you use every day. TigerBeetle is a financial ledger database built for correctness under catastrophic conditions. MotherDuck brings DuckDB - a surprisingly powerful in-process analytical database - to the cloud. Polar Signals makes continuous profiling observability practical. She is betting on the next layer of technical plumbing that serious engineers actually want to use.

There is a distinction Vais returns to often: the difference between technology and product. Engineering creates the former. Founders must create the latter. She is drawn, specifically, to the people who can do both - who understand the deep systems, and can also make them feel easy to a developer who just wants to ship something. That instinct came from watching it done poorly and well across her own career.

At Google, she was tapped for a Google.org fellowship - a program that takes Google engineers and points them at hard societal problems. Her team built a system that used computer vision and satellite imagery to track carbon emissions. TIME Magazine named it one of the 100 Best Inventions of 2020. It is a footnote in most summaries of her career. It probably should not be.

Before Spark, Vais was a Principal at Amplify Partners, where she cut her VC teeth on enterprise companies at their earliest stages. That's where her investment in MotherDuck originated - she had been tracking the thesis that analytic workloads could be run cheaply in-process using DuckDB before the rest of the market caught on. That kind of early pattern recognition, built on technical instinct rather than market hype, became her calling card.

Her path into venture was not the MBA-into-associate-into-partner route. She came through engineering, then product, then investing - each stage building on the last. USC trained her in Industrial and Systems Engineering, a discipline that is fundamentally about understanding and optimizing complex systems. Oracle gave her the hands-on experience with very large databases. The Twitter-acquired startup taught her what happens when a product grows faster than its infrastructure. Google showed her what world-class tooling feels like when it works - and when it doesn't.

At Spark Capital, she joined a firm with deep history in consumer internet - Discord, Postmates, Twitter - and brought a sharper technical infrastructure focus. When she writes about TigerBeetle, she does not reach for analogies. She explains double-entry accounting, transaction semantics, and why correctness at the database layer matters for payments at scale. She has board seats at TigerBeetle and Polar Signals.

She publishes "Natalie's notes" on Substack - a newsletter about databases, distributed systems, and the conferences worth attending in the field. It is not a marketing vehicle. It is the kind of thing you write when you genuinely care about the subject matter and want other people to care about it too. The conference guides she produces - curated lists of database and systems events for practitioners - are referenced by engineers who have no idea she's an investor.

In November 2024, she appeared on a panel at KubeCon's CNCF-hosted co-located events titled "The $100B Opportunity for the Cloud-Native Ecosystem: A VC Perspective." The framing is telling - not "opportunities I'm excited about" but an attempt to put a number on an entire ecosystem shift. Her co-panelists included investors from Unusual Ventures, Felicis, and Google Ventures. She was the one who had shipped Firestore.

Spark Capital describes her succinctly: "Natalie is beloved by her founders; she would go to the ends of the world to help them." That sentence was written to announce her partnership. It lands differently once you know the CV behind it - the engineering career, the product launches, the Google.org project. The founders she backs are not getting a generalist with a spreadsheet. They are getting someone who has sat in their chair.

Her investment thesis compresses neatly: great products share exceptional technical foundations. If the foundation is weak, the product is built on sand. She has seen it from both sides. The startups she backs are the ones building strong foundations for other people's products - the infrastructure layer that makes everyone else's code more reliable, faster, or cheaper to run. The founders she chases are the ones who can explain why their foundation matters to a developer who doesn't care about theory, just results.

The cloud-native ecosystem, developer tools, and AI infrastructure are not sleepy verticals. They are where the next generation of software gets decided. Vais is placing bets at that level - not on the apps, but on the layer underneath. That is exactly the bet you would expect from someone whose career started inside the database.