Audio / Multimodal ML Engineer (AI)

TL;DR

Audio / Multimodal ML Engineer (AI): Training and shipping speech, audio, and multimodal models for an AI safety platform with an accent on model optimization and large-scale training. Focus on building audio data pipelines, optimizing models for production via quantization and distillation, and deploying low-latency serving systems.

Location: Hybrid in Paris, France. Relocation package available.

Salary: $100,000 – $250,000 + equity

Company

hirify.global is an AI Safety company building a reliability and optimization layer for AI systems through natural-language policies.

What you will do

• Train and fine-tune large-scale audio and multimodal models from scratch or pretrained checkpoints.
• Design and execute experiments regarding architectures, data mixes, and training recipes.
• Build and maintain end-to-end audio data pipelines from raw recordings to training-ready datasets.
• Optimize models for production using quantization, distillation, and streaming inference.
• Deploy models from research checkpoints to low-latency serving environments.
• Define and implement evaluation metrics and benchmarks for product impact.

Requirements

• 3+ years of experience training large-scale deep learning models in audio, speech, or acoustic domains.
• Proficiency with PyTorch and distributed training tools like DeepSpeed or FSDP.
• Experience with audio/speech (Whisper, HuBERT) and multimodal (Audio Flamingo, Omni Qwen) architectures.
• Proven track record of shipping models to production with a focus on meeting latency targets.
• Strong knowledge of audio signal processing fundamentals including spectrograms and mel features.
• Experience with alignment techniques such as SFT, DPO, or GRPO in a multimodal setting.

Culture & Benefits

• Competitive salary and equity.
• Paid time off in accordance with local regulations.
• High-tier medical insurance in France.
• Provision of all necessary hardware, tools, and covered AI agent/IDE subscriptions.
• Bi-annual team off-sites in premium locations.

Hiring process

• Introduction call with a team member.
• Technical take-home assignment.
• In-depth technical interview.
• Final discussion with the CEO and CTO.