Founding AI/ML Engineer

VLM Run•Santa Clara, CA

1d•Onsite

About The Position

Join us as we build VLM Run – the enterprise infrastructure layer for visual intelligence. Our mission is to give developers a unified way to fine-tune, specialize, and run Vision-Language Models (VLMs) that turn images, PDFs, screenshots, and video into reliable, schema-true structured data for production insights and automation – built for scale, security, and SLAs. We’re looking for exceptional engineers to help us build and scale the infrastructure layer for visual intelligence. You’ll do well here if you bring strong technical craft, high ownership, and strength in one or more of these areas: Platform & Infra: Own and optimize the VLM inference stack (see Orion) end-to-end – from GPU serving and latency/cost to scalable backend systems and reliability. Developer Experience: Design clean, ergonomic APIs for multimodal apps – tool/function calling, structured outputs, and workflows developers actually want to build. High Agency + Velocity: We move fast on hard problems. You’ll take ideas from 0→1, set the bar for quality, and help define what “production-grade visual intelligence” looks like.

Requirements

BS & 4+ YoE
Integrated or built applications with LLMs (OpenAI, HuggingFace, Ollama, vLLM), with an understanding of prompt engineering, function calling, and structured outputs.
Python, FastAPI, async API design, schema validation, caching, and performance optimization.
Docker, Kubernetes, CI/CD, observability (logging, metrics, tracing), GCP or AWS.
Postgres, MongoDB, Redis; experience with scalable, reliable data pipelines.
Strong testing discipline (TDD), clean code, GitHub workflows (PRs, reviews, CI), and internal tooling mindset.

Nice To Haves

Shipped full-stack dev platforms or SaaS products – from landing pages to auth, billing, telemetry, and infra.

Responsibilities

Own and optimize the VLM inference stack end-to-end – from GPU serving and latency/cost to scalable backend systems and reliability.
Design clean, ergonomic APIs for multimodal apps – tool/function calling, structured outputs, and workflows developers actually want to build.
Take ideas from 0→1, set the bar for quality, and help define what “production-grade visual intelligence” looks like.