Machine Learning Engineer

The opportunity

In January 2024, a man paralyzed below the shoulders received a Neuralink implant. Within days he was playing chess and Civilization VI by imagining a cursor moving. That was our first product experience: computer control decoded from 1,024 electrodes in the brain’s motor cortex.

Since then, 20+ people have been using the device, some for 17+ hours per day.

We are now working on two problems nobody has solved.

Decoding the full virtual arm. This requires deciphering 29 degrees of freedom against millisecond timescale neural spikes. The published state of the art is four degrees of individuated finger control (Willsey et al., Nature Medicine 2025). We aim to fully decipher human intent of controlling anything that a human hand is capable of doing, and build a product that our users rely on for their independence every day.

Brain to voice. The frontier of real-time voice synthesis from intracortical signals is intelligible roughly half the time and carries only coarse pitch control (Wairagkar et al., Nature, 2025). We are working towards prosodic speech, in users’ own voices, streaming fast enough to hold a natural conversation.

If you want to help us solve these problems, we want to hear from you.

Why this is an interesting machine learning problem

• Data. Train on >50,000 hours of neural data from clinical trial participants
• Nonstationarity. Tackle the tough open problem that neural activity drifts and we need to solve decoding while minimizing user recalibration
• Co-adaptation. The brain adapts to your decoder while your decoder adapts to the brain
• Strict constraints. The brain implant operates on a tight power and the participant experience requires optimizing every millisecond of latency
• Your eval is a person. Success is a human being capable of doing something today that they couldn't do yesterday

What you'll work on

• Train models on neural spike data across sessions and participants so a new user’s decode works out of the box, and stays working through weeks of drift without supervised recalibration
• Design online adaptation that keeps the closed loop stable while both the model and the user are learning
• Compress a full hand-and-arm decoder to run with a tight latency budget on the user’s device
• Build the pre-training and post-training stack for voice synthesis: self-supervised representation learning on neural recordings, supervised fine-tuning on paired data, reinforcement learning with intelligibility and naturalness rewards, feeding a streaming vocoder that produces speech in real time
• Define what evidence justifies shipping a model update to someone’s brain-computer interface, then build the eval and rollout infrastructure to meet that bar
• Feed learnings back into the hardware platform. Our ASIC, thin-film arrays, and the rest of the brain implant are built in-house and follow your engineering gradient

You might be a good fit if you

• Have designed, trained, and shipped real-time ML systems that people depend on
• Have strong sequence modeling instincts (e.g., in speech, robotics, reinforcement learning, time-series, sensor fusion)
• Prefer owning a problem end to end (data, model, deployment, eval) over one layer of a big stack
• Are deeply curious to understand and expand human consciousness
• Bachelor’s degree in Computer Science, a related field, or equivalent experience

Helpful (not required)

• Familiarity with intracortical brain-computer interface decoding literature (e.g., handwriting and speech decoding, manifold alignment, unsupervised recalibration, cross-participant transfer)
• Experience with model compression, quantization, or inference on constrained hardware
• Experience in small-data or heavy-distribution-shift regimes

Note: No neuroscience background is required. We value simple solutions built from first principles, and some of our best decoding work has come from people who have zero experience with neuroscience.

Our tech stack

• Machine Learning: PyTorch, JAX
• Backend: Python, Rust, Swift, Bazel
• Frontend: Typescript, SwiftUI

You may not be a fit if

• You prefer remote work. This role is on-site five days a week. The lab, the robot, and your teammates are in the building. So is the job.
• You’re looking for a 9-to-5 job. The pace is intense. People work hard here because a human being is waiting on the next release.
• You prefer narrow, specific scope. We have small teams and high ownership.

Hiring process

Typically four weeks from start-to-finish.

1. Application: we ask for three concise examples of exceptional ability
2. Recruiter call: 30 minute video call with a member of our talent team
3. Hiring manager call: 45 minute technical video call with the hiring manager
4. Technical interview: 45 minute video call with an ML engineer about system design
5. On-site day: live presentation of a technical problem you’ve worked on plus a series of interviews with your future teammates
6. Founder interview: 15 minute interview with our Co-Founder, DJ Seo

Decision within 1 business day of your final interview.

Learn more