[Remote] Machine Learning Engineer.

Note: The job is a remote job and is open to candidates in USA. Runflow is a company focused on AI quality scoring for images and videos, and they are seeking a Machine Learning Engineer to enhance their scoring layer. The role involves hands-on machine learning work, pushing the scoring layer forward and extending it to new content types while ensuring quality in production environments.ResponsibilitiesYou'll own the full lifecycle, from prototype to production at scaleArchitect the evaluation pipeline, router, preprocessors, judges, for image and videoTrain, fine-tune, and quantise specialised neural networks (face, pose, segmentation, OCR, embeddings, depth)Design dynamic rubrics that adapt to use case, headshot vs. product vs. fashion vs. videoSpawn LLM judges per criterion. Decide model, prompt, complexity tierExtend Sentinel from images to video, temporal consistency, motion, identity drift, lip syncBuild the eval harness, labelled samples per use case, CI regression alertsPush the LLM router toward the cost / quality / latency frontier (Lite / Flash / Pro tiers)Co-design the auto-fix loop with workflow engineers, generate → evaluate → fix → re-evaluateSkillsYou've trained or fine-tuned a small neural network end-to-end, datasets, loss curves, the worksYou've shipped quantised inference (INT4/INT8) and know what trade-offs are acceptable in productionYou've designed an evaluation pipeline for any generative task and made it run reliablyYou have strong opinions on when to use an LLM vs. a specialised model, and can defend themYou're comfortable shipping in TypeScript even if Python is your home, the system spans bothYou read recent ML papers (LLM-as-judge, eval design, video gen) and can summarise them in 3 sentencesHave published an evaluation framework or LLM-as-judge benchmarkHave worked on AI video generation or video evaluation specificallyHave hands-on experience with Qwen3-VL, SAM 3.1, OpenPose, Gemini Embedding 2, ArcFace, or similar specialised CV / multimodal modelsHave built a model router or orchestrator for multi-model production systemsHave shipped open-source ML infrastructure or contributed to one of the libraries aboveCompany OverviewAI Image Models that actually work, with one API call. Benchmarked, production-ready image pipelines, without the complexity & cost. It was founded in 2025, and is headquartered in Dover, Delaware, US, with a workforce of 2 - 10. Its website is https://www.runflow.io. Apply To this Job