About the Company

Hushh.ai is building seamless AI and data systems. We're a small, motivated team with a flat structure, valuing initiative and excellence. Strong communication is key. Our Hushh Garages provide a unique live-work environment to foster innovation.

Why Hushh ?

At Hushh, you'll join an innovative, ambitious startup that values creativity, perseverance, and teamwork. We believe in the transformational power of personal data in the hands of the individual and are committed to building a platform that champions this belief. As part of our team, you'll have the rare opportunity to shape something meaningful from its inception.

Hushh is an equal opportunity employer, championing inclusivity and diversity. We welcome all qualified applicants irrespective of race, religion, gender, sexual orientation, age, disability, or veteran status.

Responsibilities

• Develop cutting-edge AI solutions that seamlessly integrate with user devices, focusing on personal assistants like Siri, Alexa, Microsoft Copilot, and more.
• Work on optimizing on-device AI/ML for both edge and cloud computing, ensuring cost-effectiveness and high performance.
• Collaborate within a flat organizational structure where initiative, creativity, and a consistent delivery of excellence are highly valued.
• Engage in challenging projects that span various aspects of AI and data management, from training large-scale neural networks to inventing new algorithms.
• Thrive in our dynamic Hushh Garages environment.

Basic Qualifications

• Bachelor’s or Master’s degree in Computer Science, Engineering, AI, ML, or a related field.
• Strong passion for solving complex problems in AI and data systems.
• Ability to work hands-on and contribute directly to the company’s mission.
• Excellent communication skills, with the ability to share knowledge concisely and accurately.

Additional Requirements

• Experience building distributed training systems for AI models in HPC clusters.
• Expertise in JAX/XLA for multi-host training.
• Proficiency in writing custom CUDA kernels and optimizing GPU utilization.