Logical Intelligence · San Francisco, CA

Salary
Posted
Jul 25, 2026
Location
San Francisco, CA
Last confirmed open
Jul 26, 2026

What this job asks for AI summary

This role centers on building and improving LLM pipelines for distributed training, implementing novel reasoning algorithms, and modifying state-of-the-art model architectures. The work spans pre-training, fine-tuning, and optimization of large language models, with a focus on formal verification and logical reasoning applications. It suits engineers with hands-on ML infrastructure experience, strong knowledge of transformer internals, and comfort working across Python, C++, and deep learning frameworks.

Mid level · 3+ years

Must have (6)
LLMsPython or C++PyTorch, TensorFlow or JaxKubernetesdistributed training · 3+ yrstransformers
Nice to have (2)
energy-based modelslatent space reasoning

“or” means any one of them counts — you don't need all of them.

We read this from the posting text with AI. Skim the description below before ruling yourself out.

How this req sits in the market our data

What gives you an edge
PyTorch9%transformers9%

Rare in this occupation — lead with these, and say what you built with them.

What won't set you apart
Python51%

Most people in this occupation already list these. Still required — just not what gets you shortlisted.

What the occupation pays Median $138,970 (middle half $107,524–$175,762).

Estimated from BLS employment for this occupation and area, per-skill prevalence across our listing corpus, and published wage benchmarks — as of Jul 28, 2026. It is a model, not a headcount.

Why we read it this way (5)

The role title is 'AI Engineer' with no seniority modifier; advertised seniority is Unspecified. The 3+ years requirement and scope (implementing/optimizing LLM pipelines, modifying architectures) support a Mid-level classification.

SOC is a genuine toss-up between 15-1252 (Software Developers — building and shipping LLM pipelines and training infrastructure) and 15-2051 (Data Scientists — ML modeling, pre-training, fine-tuning). The primary day-to-day work leans toward engineering and shipping systems, so 15-1252 is chosen with 15-2051 as the runner-up.

Python and C++ are listed together as a single requirement ('Python/C++'); Python is used as the primary name with C++ in alternatives. PyTorch, TensorFlow, and JAX are similarly listed as interchangeable deep learning framework options.

Energy-based models (EBM) and latent space reasoning appear under 'Bonus Points' and are marked preferred accordingly.

No compensation, employment type, or degree requirement is stated in the posting.

Read the full posting

The employer publishes the full description on their own site — read it there ↗. Or sign in to read it here — it's free, and it also lets you track this application.

Apply

Apply on employer site ↗