Multimodal AI Engineer — Media, Speech & Audio

Meta; Little Rock (AR); On-site; USD 154,000 - 217,000; Full time

Job summary

Meta is seeking a Software Engineer for Multimedia & Multimodal AI within Applied AI (AAI). You will own data pipelines and evaluation frameworks across image, video, audio, and speech modalities, building scalable workflows and agent-driven data production.

You will mentor engineers and help advance state‑of‑the‑art multimodal AI with rigorous ablations and responsible practices. You will work across model inputs and outputs, ensuring high-quality data, robust pipelines, and reliable training.

Qualifications

• 3+ years building ML systems in production or research settings
• Strong Python and PyTorch experience
• Demonstrated experience with speech, audio, or music ML

Responsibilities

• Design and build agentic workflows and pipelines, including human-in-the-loop designs
• Design and own data pipelines at scale: ingestion, filtering, pseudo-labeling and captioning
• Build evaluation infrastructure: objective metrics and human listening-test pipelines
• Improve training efficiency and reliability — distributed training and GPU utilization
• Reproduce and extend state-of-the-art research by implementing new methods from papers

Skills

Python

PyTorch

ML systems

Distributed training

Multimodal ML

Job description

Meta is seeking a Software Engineer for Multimedia & Multimodal AI within Applied AI (AAI). You will own data pipelines and evaluation frameworks across image, video, audio, and speech modalities, building scalable workflows and agent-driven data production.

You will mentor engineers and help advance state‑of‑the‑art multimodal AI with rigorous ablations and responsible practices. You will work across model inputs and outputs, ensuring high-quality data, robust pipelines, and reliable training.