Senior Machine Learning Engineer, Grounded AI

Moraph LLC; San Francisco (CA); On-site; USD 285,000 - 325,000; Full time

Job summary

Moraph LLC is building AI that regulators can trust, developing retrieval-augmented models for FOIA redaction and citizen-services products. You will design pipelines that answer only from approved client sources and ensure all outputs are Grounded and explainable.

We seek engineers with 5+ years of experience deploying ML systems, hands-on work with LLMs and vector databases, and strong Python skills. You will join a mission-driven team ensuring data stays within client boundaries across

Qualifications

• 5+ years building and deploying machine learning systems in production.
• Hands-on experience with LLMs, embeddings and retrieval systems.
• Strong Python and familiarity with modern ML tooling (PyTorch, vector databases, model serving).
• A habit of measuring model quality rigorously rather than by demo.

Responsibilities

• Design and ship retrieval-augmented generation pipelines that only answer from approved client sources.
• Build evaluation harnesses that measure grounding, redaction accuracy and hallucination rates before release.
• Fine-tune and serve models for PII and CUI detection in high-volume document workflows.
• Partner with the Model Risk team so every production model has documentation, monitoring and a rollback plan.
• Keep sensitive data inside client boundaries across private and public cloud.

Skills

Production ML systems

LLMs retrieval systems

Python programming

Model evaluation

Tools

PyTorch

Vector databases

Model serving

Job description

Build AI that regulators can trust. You will develop the retrieval-augmented models behind our FOIA redaction and citizen-services products, where every answer must be grounded in a client's own verified documents and every decision must be explainable.

What you'll do

• Design and ship retrieval-augmented generation pipelines that only answer from approved client sources
• Build evaluation harnesses that measure grounding, redaction accuracy and hallucination rates before release
• Fine-tune and serve models for PII and CUI detection in high-volume document workflows
• Partner with the Model Risk team so every production model has documentation, monitoring and a rollback plan
• Keep sensitive data inside client boundaries across private and public cloud

What you'll bring

• 5+ years building and deploying machine learning systems in production
• Hands-on experience with LLMs, embeddings and retrieval systems
• Strong Python and familiarity with modern ML tooling (PyTorch, vector databases, model serving)
• A habit of measuring model quality rigorously rather than by demo

Compensation

$285,000 – $325,000 per year.