About Us
Nuha is a research lab based in New York City. We teach companions how to understand, interact with, and respond to humans and the physical world.
Your Role
You'll build the models that give our companions perception and understanding. You'll work at the intersection of ML research and product. Your models will ship to real users and help create magical interaction experiences.
What You'll Work On
Multimodal Model Development
- Design architectures that fuse multiple sensor modalities into unified representations
- Build models that understand spatial context, user presence, and environmental dynamics
- Develop systems for continuous learning and adaptation based on user interactions
- Create lightweight inference pipelines that run efficiently on our hardware constraints
Research Problems
- How do we make companions spatially aware and context-sensitive?
- How do we build memory systems that feel natural and respectful?
- How do we personalize interactions while maintaining privacy?
- How do we bridge the gap between raw sensor data and meaningful understanding?
- How do we run large models locally on small devices?
Ship & Iterate
- Train and deploy models to production and measure real-world performance
- Work directly with users to understand what makes companions feel wonderful
- Balance research innovation with shipping features that delight users
Requirements
- Strong ML fundamentals - You can design architectures, debug training, and ship models to production
- Multimodal thinking - Experience working with systems that combine different data types (audio, vision, sensor, text)
- Research mindset with execution bias - You publish when it matters, but you ship products
- Python/PyTorch - You write clean, efficient model code
- First principles reasoning - You can break down ambiguous problems and figure out what to build
Bonus
- Mathematical maturity
- Signal processing background (radar, RF, audio, ultrasound)
- Vision systems experience (SLAM, depth estimation, 3D reconstruction, tracking)
- On-device ML / model optimization
- C/C++
- Robotics or embodied AI research
- Publishing record in relevant conferences (CVPR, ICRA, NeurIPS, etc.)
Team
You'll work alongside engineers doing: RF/mmWave radar design, PCB/enclosure design, embedded client software, backend infrastructure, data pipelines, and manufacturing systems.
Interview Process
- Phone screen
- Technical deep dive - Walk us through a research project you've owned end-to-end
- In-person whiteboard - Design a solution for a specific scenario
- Work trial - Spend time with the team prototyping a real model or system
Location
NYC - In office. We're building a physical product and need to work together in person.
Apply
Send an email to: careers@nuha.com
Include:
- Resume + GitHub, HuggingFace, Kaggle, etc.
- What excites you about this role
- A research problem or project you've owned