Deterministic ML Research Scientist: Reproducible AI

Serve AI; Kansas City (MO); On-site; USD 120,000 - 170,000; Full time

Job summary

Serve AI is building deterministic enterprise AI infrastructure focused on grounded, auditable results. The role centers on research to push accuracy within a fully reproducible and explainable framework, ensuring every step can be audited and validated.

You will design experiments, reproduce claims, and report findings with precision, then hand off validated methods to the engineering team for production use.

Qualifications

• Experience researching embeddings, representation learning, or metric learning under architectural constraints.
• Ability to design and reproduce experiments with precise detail and documentation.
• Strong written communication for reporting methods and results.

Responsibilities

• Design and run experiments to improve routing, grounding, or composition while preserving deterministic output.
• Reproduce and validate externally claimed results before they influence roadmaps.
• Document negative and inconclusive results with sufficient detail for replication.
• Hand off validated methods to engineers for production integration.

Skills

Embeddings

Metric learning

Representation learning

Deterministic ML

Statistical power

Technical writing

Auditable results

Experiment design

Tools

Python

PyTorch

TensorFlow

Job description

Serve AI is building deterministic enterprise AI infrastructure focused on grounded, auditable results. The role centers on research to push accuracy within a fully reproducible and explainable framework, ensuring every step can be audited and validated.

You will design experiments, reproduce claims, and report findings with precision, then hand off validated methods to the engineering team for production use.