Talent Map Brief: ML Platform & LLM Infrastructure in the Bay Area
A worked market hypothesis for ML platform, inference, model-serving, and distributed-systems talent in the San Francisco Bay Area.
Role: ML Platform / LLM Infrastructure Engineer · Geography: San Francisco Bay Area · Last reviewed: 2026-09-06
Ask SourcingOS about this page
Literal role requirements
- Production ML/platform engineering scope defined by the role
- Required distributed-systems or inference experience only where explicit
- Location/hybrid expectations
- Seniority and ownership expectations
Title families to test
- ML Platform Engineer
- Machine Learning Infrastructure Engineer
- ML Systems Engineer
- Inference Engineer
- Distributed Systems Engineer
- AI Infrastructure Engineer
- Platform Engineer
Donor categories
- Frontier-model and AI infrastructure companies
- Cloud and compute platforms
- Large-scale consumer/product companies with internal ML platforms
- Developer infrastructure companies
- Data/streaming/vector/search infrastructure companies
Geography bands
- San Francisco
- Peninsula / South San Francisco / San Mateo
- Palo Alto / Mountain View / Sunnyvale
- San Jose / Santa Clara
- East Bay where commute/hybrid expectations are viable
Evidence and source lanes
- Direct-title search
- GitHub repositories and technical contributions
- Research/publication evidence for systems-heavy candidates
- Open-source serving/inference ecosystem evidence
- Conference/talk/podcast technical evidence
- ATS rediscovery
- Licensed people-data expansion
Questions that determine whether the market is actually scarce
- Does the role need research depth or production systems depth?
- Is GPU/CUDA experience required or simply adjacent?
- Which inference/runtime technologies are truly mandatory?
- How much scale/latency evidence is needed?
- Is Bay Area presence required or is remote talent acceptable?
Actionable search plan
- Build separate research, open-source, and production-platform lanes.
- Do not use “AI engineer” as a catch-all title family.
- Distinguish model-building experience from serving/infrastructure ownership.
- Use public evidence to deepen capability confidence without assuming current employer scope.
- Calibrate on the first slate around systems depth, model proximity, and production ownership.