Davis Treybig

Background

• I grew up in Austin, Texas. As a kid, I was very interested in computers, math, and science. I also generally loved to tinker with things - I was notorious for dissasembling things in the house like electrical fixtures, and for breaking into my dad's toolshed to steal tools. I once took a saw and cut a giant hole into our garage.
• I was lucky to get a lot of exposure to entrepreneurship and startups at a young age. My dad worked with a number of founders and would often talk to me about technology startups. In High School, he frequently let me sit in on sessions where he would work with founders on pitches or business strategy. This is what drove my early interest in technology startups.
• I studied Computer Science & Electrical Engineering in college at Duke University. College is also where I developed more of an interest in design tools - I taught myself how to use Sketch & Photoshop for a few clubs I was part of. Digital design tools were a revelation for me, as growing up I never considered myself artistic or creative because I was really bad at drawing & painting. The more mathematical nature of digital design tools deeply appealed to me.
• I loved studying computer science and initially figured that I should be a software engineer, but after interning as one I realized I didn't like the day to day as much. I then tried product management and realized it was a much better fit - I still got to work closely with engineers on interesting technical topics, but it was also much more interdisciplinary.
• After school, I worked at Google for ~3.5 years as a product manager. I worked on the early developer platform for the Google Home, where we tried to figure out how to let people design conversational applications before LLMs (hint - it turned out to be really, really hard...). I then worked on the Pixel Phone team. Two of the cooler Pixel features I worked on were 1. Launching Google's second-ever federated learning feature, a federated learning-based model for search in Settings and 2. Working with a Brain team on utilizing a predictive battery model for various battery-related features on the phone.
• Google was a great place to be out of college, but I'd always been far more interested in startups. A messy acquisiton of HTC gave me a crash course on how hard it can be to merge two companies with such profound cultural differences, and led me to decide it was time to move on. I had loved working on developer tools, I was always one of the most data-driven PMs on the teams I was on, so I was mostly focused on developer tooling & data infrastructure startups. But, I coincidentally had a venture recruiter reach out around this time, and decided I might as well interview.
• I've now been in venture for ~6 years, and I absolutely love it. What drew me to venture was a love of learning, a sincere interest in technology & science, and a deep admiration of the startup journey.

Investing

Companies I have invested in & work with as part of my job at Innovation Endeavors

• Eppo - Modern experimentation platform. Led Series B.
• Augment - Copilot for software engineers. Co-Led Series B.
• Dosu - AI Support Engineer. Led Pre-seed and Seed.
• Unnanounced company rethinking motion graphics w/ WebGPU & Diffusion models. Led A.
• Sequence - Browser-based, AI native professional-grade video editor. Led Seed.
• Responsive - Stateless Kafka Streams for reactive backends. Led Seed.
• Bauplan - Serverless compute runtime for SQL + Python workloads on data lakes. Led Seed.
• Earthly - Modern build platform for fast, reproducible builds. Led Seed.
• Continual - AI Copilot as a service. Led A.
• Pyte - Cryptograpy-based data collaboration on private data. Led Seed.
• Tromzo - Automated triage & remediation for application security teams. Led pre-seed.
• Authomize - Cloud identity & access management security. Led Series A. Acquired by Delinea
• Panther - Snowflake-native logs analysis for security engineering teams. Led Seed.
• InsidePacket - Programmable network switch allowing for embedded load balancing, DDOS protection, and other programmable network logic. Led Seed
• Viaduct - Timeseries forecasting for predictive maintenance and root cause analysis of massive-scale sensor data. Led Series A .

Writing

I really like to write. I am a big believer in writing as a method for distilling thought and for ensuring you really understand something. I also believe in the idea of writing as a conversation. A few of my favorite things I've written over the years:

• S3 as the Universal Infrastructure Backend - How databases are being redesigned around blob stores as their persistent storage layer
• The Experimentation Gap - An overview of how A/B testing and experimentation has changed from a technical & market perspective.
• Software Systems in a World of LLMs - Why many software systems (languages, APIs, SDKs, etc) will likely need to be rebuilt for AI consumers rather than humans
• WebGPU: What is it and why does it matter? - A technical overview of WebGPU and its likely path to becoming the standard GPU interface beyond the browser
• The Biggest Bottleneck for Large Language Model Startups is UX - A discussion on how UX design is an under-appreciated challenge in building foundation model-based applications
• The Context Window Dillema - An overview of different research approaches to expanding context window sizes
• Non-destructive Editing in Design Tools- A technical lens on why non-destructive editing is becoming a more prominent paradigm in design tools thanks to computing advances, and it's implications
• The Analytical Application Stack - How data infrastucture needs for customer-facing and "operational" data systems differ from classical data infra
• Are Foundation Model Applications Complex Systems? - An exploration of how foundation model based applications need to be thought of as complex systems from an evaluation & system design perspective
• Generative AI and the Shift from Creation to Validation - How advances in AI are suddenly making evaluation the bottleneck of most systems

Other things about me

• I love gardening and cooking, especially cooking things I grow myself! I have a small garden of 25+ plants and trees, and I am probably a bit too addicted to Youtube cooking videos. I originally got into cooking from the show "Good Eats", a great show which focused on the science of cooking and food. My biggest cooking flaw is I can never just follow a recipe - I always have to modify or change it in some way.
• I really enjoy weightlifting, and have lifted consistently since High School. My best ever lifts are a 425 lb deadlift, a 305 lb squat, and a 255 lb bench. Recently, I have gotten very into kettlebell workouts.
• There was a period of time where I was one of the most read writers about "Associate Product Manager" programs on Quora (back when Quora was a thing). This led to a friend of mine and I self-publishing a book on breaking into APM programs out of college - The Product Diploma.
• For a period of time, some college friends & I maintained a blog we called Five Guys Fun Facts where we each wrote about one interesting fact we learned a week. I am biased but I think it is legitimately interesting content!
• As you can probbaly tell from my investments, I am interested in data analysis. One of my earliest forays in this area was figuring how to analyze my core college friend groups' iMessage group text data when I was in college. You can find the original analysis here.
• Other things I love: Sparkling water. "Generative art" before the AI craze - using math & randomness to generate interesting abstract patterns like this via tools like Processing. Landscape photography - I don't do it myself, but I find it beautiful. Documentaries - some of my favorites are Sour Grapes, Icarus, and Meru.

Reading