Founding Lead Machine Learning Engineer
Ist diese Stelle etwas für Sie?
Lebenslauf erstellen Erstellen Sie Ihren Lebenslauf und entdecken Sie Ihre Übereinstimmung mit dieser Stelle — und mit allen anderen.
Bewerbungen unterwegs verfolgen Die kostenlose Whileresume-App für iPhone und Android.
Die Stelle
You will lead the technical direction for the first product, defining model selection, training strategy, infrastructure, and long-term architecture.
Own end-to-end ML systems—from data pipelines and training to evaluation and scalable inference deployment.
Design or adapt frontier models (LLaMA, Mistral, Qwen, Claude-compatible) and pioneer approaches where existing ones fall short.
Fine-tune models using state-of-the-art methods (LoRA/QLoRA, SFT, DPO, distillation) and architect scalable inference (vLLM, TensorRT-LLM, DeepSpeed).
Build data systems for high-quality synthetic and real-world training data; develop alignment, safety, and guardrails; design robust evaluation frameworks.
Shape early product direction as a founding builder—own the problems end-to-end, move quickly with discipline, and influence the product's long-term direction.
Own end-to-end ML systems—from data pipelines and training to evaluation and scalable inference deployment.
Design or adapt frontier models (LLaMA, Mistral, Qwen, Claude-compatible) and pioneer approaches where existing ones fall short.
Fine-tune models using state-of-the-art methods (LoRA/QLoRA, SFT, DPO, distillation) and architect scalable inference (vLLM, TensorRT-LLM, DeepSpeed).
Build data systems for high-quality synthetic and real-world training data; develop alignment, safety, and guardrails; design robust evaluation frameworks.
Shape early product direction as a founding builder—own the problems end-to-end, move quickly with discipline, and influence the product's long-term direction.
Die vollständige Anzeige sehen
Aufgaben, Anforderungen, Kompetenzen und Vorteile — mit Ihrem kostenlosen Konto.
oder
Bereits ein Konto?
AnmeldenDiese Stellen könnten Sie interessieren
Noch keine wirklich ähnliche Stelle — hier die neuesten.