Senior Machine Learning Engineer / Research Scientist

Tools for Humanity (TFH) designs and builds technology behind World. World is building a real human network designed to accelerate people in the age of AI. As bots and autonomous agents reshape the internet, people, institutions, and applications need a trusted way to confirm who is a real human while preserving privacy. The TFH and World tech stacks make this possible: the Orb verifies real, unique people, World ID proves it privately, and World App puts these capabilities, and more, in people’s hands. Together, they add a human layer to an AI-driven internet. World is already running at a global scale. More than 17 million people across 160 countries have verified with World ID, and more new Orb verifications take place each week. World App is already among the most used wallets globally. Developers are integrating World ID to build safer online experiences and create spaces where real people can participate, earn, and be recognized in ways AI simply can’t replicate. Founded in 2019, TFH has more than 400 people across hardware, software, AI, cryptography, mobile engineering, and global operations. Our teams come from OpenAI, Tesla, SpaceX, Apple, Google, Stripe, Meta, Coinbase, Palantir and MIT Media Lab. We’re backed by leading investors, including a16z, Khosla Ventures, Bain Capital Crypto, Blockchain Capital, Variant, Tiger Global, and Coinbase Ventures, as well as prominent operators and founders across fintech and AI. TFH and World have been featured on the cover of [TIME Magazine](https://time.com/7288387/sam-altman-orb-tools-for-humanity/), highlighted in [Fast Company’s](https://www.fastcompany.com/91411606/fintech-blockchain-next-big-things-in-tech-2025) Next 5 in Fintech, and explored in a [Bloomberg deep dive](https://www.bloomberg.com/news/features/2024-08-12/how-worldcoin-is-building-digital-ids-to-combat-the-ai-apocalypse). Our leadership is also named to the [Time AI 100](https://time.com/collections/time100-ai-2025/). Learn more about the newest product launches from our [Liftoff](https://world.org/liftoff) event.

• •Improve our core biometric identification and anti-spoofing models, training and iterating on deep learning architectures, losses, and data pipelines, with model size, latency, and memory budgets as first-class design constraints from day one.
• •Reach for classical computer vision and image processing wherever it is the right tool, whether as the actual solution to an identification, detection, or quality-assessment problem, as the preprocessing stage of a deep learning pipeline, or as diagnostic tooling for understanding what a model is seeing.
• •Lead independent research initiatives end-to-end: form a written hypothesis, design ablations that isolate variables, run experiments, read results honestly, and know when to ship and when to stop chasing the last percentage point.
• •Look at the data when models fail. Pull up misclassified samples, form concrete hypotheses about why they failed, and use that to drive the next iteration, rather than waiting for an aggregate metric to explain the problem for you.
• •Build evaluation and monitoring pipelines that catch model regressions before they reach production and surface data drift in the wild, including across the AMPC database, not weeks later in a post-mortem.
• •Take work all the way from prototype through rigorous testing to deployed system, partnering with different teams. Translate between ML, embedded, and secure-compute constraints when the conversation needs it.
• •Write design docs, experiment write-ups, and technical proposals that hold up to scrutiny across teams and remain useful months after they were written. Drive alignment on contentious decisions.
• •Help shape technical standards across the AI & Biometrics team - evaluation methodology, experimentation discipline, model versioning, monitoring, and mentor more junior researchers and engineers as a default behavior, not as an extra task.

• •An "in-the-driver's-seat" operating style: you take ownership of problems end-to-end, drive your work forward without waiting for direction, and stand behind your decisions once they are in production.
• •Strong fundamentals in classical computer vision and image processing - OpenCV, NumPy, the standard toolkit of filters, transforms, morphology, geometric methods.
• •Deep, hands-on experience training and shipping deep learning models for computer vision at production quality, including under tight latency and memory constraints. Exposure to real edge or embedded deployment is a strong plus.
• •A pragmatic, applied-research mindset: you care about rigor and depth, but you know when a result is good enough to ship and when chasing the last percentage point on a benchmark is the wrong use of your time.
• •Solid mathematical fluency at the level where you can spot pathologies in proposed designs without running them.
• •Experimental discipline: hypotheses written down before code is, ablations that isolate one variable at a time, and the judgment to know when a result is real and when it needs more seeds.
• •A collaborative operating style: you mentor, share knowledge by default, and engage constructively with constraints from neighboring teams rather than treating them as obstacles. We are not looking for a lone wolf, however brilliant.
• •Strong plus: direct experience with biometric identification at scale; margin-based metric learning losses (ArcFace, Triplet and their variants) and their failure modes; anti-spoofing / presentation attack detection; red-team or adversarial evaluation of ML systems; publications at top ML venues.