Senior Software Engineer, AI Platform and Enablement

Descript Descript · AI Frontier · Remote · Engineering

Senior Software Engineer on the AI Platform and Enablement team at Descript, focusing on integrating and managing AI models (both first-party and third-party) into a next-generation AI-powered audio and video content creation platform. Responsibilities include building and maintaining model integrations, designing and implementing AI infrastructure (data pipelines, training, evaluation, deployment), optimizing inference, and collaborating with research and product teams.

What you'd actually do

  1. Build, maintain, and standardize third-party model integrations, including consulting for other engineering teams with AI model integration needs
  2. Design, implement, and maintain our AI infrastructure supporting our machine learning life cycle, including data ingestion pipelines, training developer experience and infrastructure, evaluation frameworks, and deployments / GPU infrastructure
  3. Collaborate with Product Managers, Research Engineers, and AI Researchers to understand their infrastructure needs and ensure our AI systems are robust, scalable, and efficient
  4. Optimize and scale our models and algorithms for efficient inference
  5. Deploy, monitor, and manage AI models in production

Skills

Required

  • Deploying and managing AI models in production
  • Large volume data pipelines (spark, flume, dask)
  • Cloud platforms (AWS, Google Cloud, Azure)
  • Container technologies (Docker, Kubernetes)
  • DevOps and MLOps best practices
  • Python
  • C/C++
  • CUDA
  • GPU performance profiling
  • Distributed training
  • Machine learning frameworks (PyTorch, TensorFlow)

Nice to have

  • Generative AI models
  • Audio and video processing

What the JD emphasized

  • Experience in deploying and managing AI models in production
  • Experience with the tools of large volume data pipelines like spark, flume, dask, etc.
  • Familiarity with cloud platforms (AWS, Google Cloud, Azure) and container technologies (Docker, Kubernetes).
  • Knowledge of DevOps and MLOps best practices
  • Experience with generative AI models
  • Familiarity with audio and video processing
  • Knowledge of Python, C/C++, CUDA, and experience profiling GPU performance and distributed training runs
  • Experience with machine learning frameworks like PyTorch, TensorFlow or similar

Other signals

  • integrating first-party and third-party models
  • AI infrastructure
  • deploying and managing AI models in production
  • optimizing and scaling models for inference