You should join koodos labs for the people and our mission.

We aspire to build the best team of the 2020s. Just like PayPal in the 90s, Google in the 00s, and Stripe in the 10s, we want to be known as “a place where it’s good to be from.” If you join us, we promise to be the best place to grow your career — with the best people you’ve ever worked with. We’re also working on a very bold mission.

We care deeply about helping people connect more deeply with themselves and others. And in the spirit of personal empowerment, we aim to invert the internet’s data model over time and put individuals in control of their digital identities.

We’re the Context Company. Context is what turns generic AI into personalized intelligence. It’s all the signals available to a product: environmental cues, device state, user history, intent. Context is what transforms a cold machine into something that knows and understands you.

koodos is building Shelf, a user-controlled data store used several times a week by millions of people to track what they’re consuming and keep up with what others are into. It connects to any platform, aggregates your credentials and consumption data, learns from your activity, and, eventually, will let you share that context with services for more personalized experiences. koodos labs lets you store your digital life in one place — then take it anywhere.

The future is personalized experiences powered by user-controlled context. We’re building a future where apps and services come to you and your memory—not the other way around. See our latest paper here.

Personalization effects are the new network effects. The better an experience understands and adapts to you, the more irreplaceable it becomes. We need someone who can make that adaptation happen in milliseconds, not minutes. You’ll build the inference and personalization systems that turn rich context into magical experiences — making every interaction feel impossibly personal, instantly.

In this role you will:

We’re early stage, so you’ll have an outsized equity stake and an unparalleled opportunity to define how AI serves humanity. Engineers at koodos labs decide what gets built and why, in addition to figuring out how.

We’re NY-based (we have our own office in West Soho) and we work in-person.

We’re a small, but mighty team with backgrounds at YouTube, Coinbase, Harvard, and Cambridge, as startup founders and as early members at companies like Improbable and Lyft. We’ve come together around a shared vision and are dedicated to creating important and positive experiences for cyberspace.

We’re well resourced (unannounced rounds) & backed by top-tier investors, including the backers of companies like Airbnb, Pinterest, Snap, and Twitter, as well as the founders of companies like Zynga, VSCO, and Scale and the people behind artists like Miley Cyrus, Justin Bieber, Lorde, Logic, and Panic! At The Disco, and many others. Our team is advised by the founders of Pinterest, Dubsmash (now Reddit), as well as pioneers of digital marketing and market design from Harvard.

We care about building a genuinely diverse team. We are a majority-first gen immigrant team and sponsor visas — we think that’s important as we build towards enabling easier digital migration. We share the same values of individuality, empathy, reliability, kindness and humility. One big overlap among our life experiences is contrasts: contrasts between our own upbringing and the world around us, contrasts between what was expected of us and what we ended up pursuing and our bringing together of contrasting, interdisciplinary worlds.

More about us here.

We’re looking for someone who can bridge cutting-edge ML with real-world impact. You might be a great fit if you are:

Day 1: You’ll merge your first PR. We want our developer experience to be as smooth as possible, so this first day is a good test of how we’re doing there.

Day 7: You’ll ship your first inference optimization to production. Maybe you’ll cut latency by 30% with better caching, or implement request batching that doubles throughput. You’ll experience our full deployment pipeline and see real users benefit from your work.

Day 50: You’ll have shipped personalization that makes users say “how did it know?” You’ll have built the first version of our real-time recommendation system, integrated with our memory architecture, and created an API that third-party developers are excited to use. You’ll have established strong collaboration patterns with our Memory Engineer, shipping features that showcase intelligent context in action.

In the Future: You’ll define how the world experiences personalized AI. As our Memory Engineer expands what we can remember about users (with their permission), you’ll ensure that knowledge translates into delightful, instant experiences. You’ll build the inference infrastructure that makes “context liquidity” real — where any app can instantly understand and serve users perfectly. Your systems will prove that personalization effects are indeed the new network effects.

We aim to move fast — typically two weeks end-to-end. We find this works best for both us and you!

And we focus on real-world experience, demonstrated ability, and references. No riddles or binary tree puzzles — just a thoughtful look at what you’ve done and where you’re headed.

Introduction: We’ll kick things off with a call with one of our co-founders.

Technical Screening: A conversation with our CTO to dig into your experience with ML systems, inference optimization, and personalization. We’ll discuss your thoughts on the evolution from traditional RecSys to context-aware AI.

Take-Home Exercise: A short presentation (video or written) on either: (1) a personalization system you’ve built and its impact, (2) your approach to serving LLMs at scale with <100ms latency, or (3) how you’d design instant personalization APIs for third-party developers.

Onsite: You’ll meet the team. We spread this across two days to reduce load:

References: We’ll ask for 2–3 references, and may also reach out to others you’ve worked with (let us know if there are any sensitivities). We’ll keep it brief and respectful — usually ~15 mins per call. You’re also welcome to reference check us.

Decision: We’ll move quickly on our end.