Multimodal ML Engineer

Npv · Paris$8,333–$20,833 USD/messenior

In short

▸Ingeniero de ML multimodal que entrena y despliega modelos grandes desde cero.
▸Día a día: construye pipelines de datos, entrena arquitecturas MoE y optimiza modelos para producción con enfoque en alineación.
▸Destacado: trabajas en una startup de seguridad de IA con backing de líderes de OpenAI, Anthropic, Mistral, etc.

English Speaking Jobs in France

Apply on company site ↗Share on WhatsApp

In ~1 minute you get: who interviews you, the likely questions answered from your CV, and your CV tailored to this job. Free, no card.

🎧Land the interview? Bring the copilot. Our free extension listens to the live interview and flashes 3-4-word anchors from your resume and prep — glance, connect, talk. Get the extension →

What they ask for

✓3+ años entrenando modelos multimodales a gran escala.
✓Experiencia sólida en PyTorch y entrenamiento distribuido (DeepSpeed, FSDP).
✓Conocimiento profundo de arquitecturas multimodales (LLaVA, Qwen-VL, Audio Flamingo, Whisper, etc.).
✓Experiencia práctica en RLHF/alignamiento (DPO, GRPO, modelado de recompensas).
✓Experiencia con audio y video: modelado de secuencias, curación de datos a gran escala, optimización de inferencia.
✓Disponibilidad para mudarse a París o Londres (híbrido).

Don't tick every box? That's normal — your free dossier shows your gaps and how to cover them in the interview.

PyTorchDeepSpeedFSDPLLaVAQwen-VLInternVLAudio FlamingoWhisperHuBERTConformer

Who should you write to at Npv?

Your free dossier identifies the people who'd interview you — their background, what they value, and how to reach out so you stand out before applying.

We're looking for a Multimodal ML Engineer to join White Circle , an AI Safety company building the policy enforcement and optimization layer for AI systems. Backed by $11M from senior leaders at OpenAI, Anthropic, HuggingFace, Mistral, and DeepMind, White Circle processes 100M+ API calls monthly and runs its own LLMs in production. You will Train and fine-tune large-scale multimodal models (vision-language, audio, speech, video) from scratch and from pretrained checkpoints. Design experiments, build multimodal data pipelines, and train MoE architectures. Build alignment pipelines (SFT, DPO, GRPO), optimize for production (quantization, distillation, streaming), and deploy end-to-end. Define evaluation metrics that actually matter for the product. Requirements 3+ years training large-scale multimodal models. Strong PyTorch and distributed training experience (DeepSpeed, FSDP). Deep familiarity with multimodal architectures – LLaVA, Qwen-VL, InternVL, Audio Flamingo, Whisper, HuBERT, Conformer or similar. Hands-on RLHF/alignment across modalities (GRPO, DPO, reward modeling). Both audio and video experience required – sequence modeling for each, plus large-scale dataset curation and production inference optimization. Relocation to Paris or London (hybrid) required. Bonus Audio signal processing fundamentals – spectrograms, mel features, noise reduction. MoE architecture experience. We offer $100k–$250k/year salary + equity; higher figures can be negotiated. Official employment, visa and relocation help. Compensation: $100K – $250K • Higher figures and equity are negotiable • $100K – $250K • Higher figures and equity are negotiable Find more English Speaking Jobs in France on Arbeitnow

Don't apply unprepared

We research who's interviewing you, tailor your CV and rehearse you live — first one free.