In-Person ML Research Intern — Shape Voice AI

Phonic; San Francisco (CA); On-site; USD 45,000 - 65,000; Full time

Benefits offered by this job

Top-tier compensation

Free meals

Off-site & team events

Job summary

Phonic in San Francisco is seeking a Research Intern to tackle core problems that make voice AI feel human. You will own a research direction from problem identification to experiments, developing methods, and driving production impact with a focus on rigor and originality.

You will collaborate with researchers and engineers, building training pipelines and tooling. A track record of original work and proficiency in PyTorch or JAX are highly valued.

Qualifications

• Proven track record of original work.
• Proficiency in PyTorch or JAX.
• Strong math, probability, optimization, and linear algebra.

Responsibilities

• Identify high-leverage research problems across the voice AI stack and own them end-to-end.
• Design and run rigorous experiments to analyze architectural trade-offs for scalability, latency, and quality.
• Curate large training datasets and assess data quality impact on model behavior.
• Collaborate with research scientists and engineers to move prototypes to production.
• Build training pipelines, evaluation frameworks, and tooling for rapid iteration.

Skills

Original research

PyTorch/JAX

Mathematics

Communication

Implementation from papers

Tools

PyTorch

JAX

Job description

Phonic in San Francisco is seeking a Research Intern to tackle core problems that make voice AI feel human. You will own a research direction from problem identification to experiments, developing methods, and driving production impact with a focus on rigor and originality.

You will collaborate with researchers and engineers, building training pipelines and tooling. A track record of original work and proficiency in PyTorch or JAX are highly valued.