Senior Software Engineer, AI Coding Tools

ByteDance ByteDance · Big Tech · San Jose, CA · R&D

Senior Software Engineer, AI Coding Tools at ByteDance, focusing on the TRAE product, an AI software development agent. The role involves designing, training, fine-tuning, optimizing, quantizing, and deploying LLMs for code generation and reasoning capabilities. It requires experience with large-scale distributed training, model optimization techniques, and deployment on GPU clusters.

What you'd actually do

  1. Design, train, and fine-tune large language models (LLMs) that support TRAE’s core reasoning and code generation capabilities.
  2. Build efficient, stable, and scalable model training and evaluation pipelines.
  3. Collaborate with infrastructure and product teams to deploy and monitor models efficiently on GPU clusters.
  4. Continuously optimize models for latency, throughput, and accuracy.
  5. Stay up to date with and apply cutting-edge techniques in large model optimization and inference acceleration.

Skills

Required

  • PyTorch or TensorFlow
  • Python
  • C++/CUDA
  • large-scale distributed training
  • mixed-precision training
  • model parallelism
  • model quantization
  • pruning
  • CUDA-based deployment optimization

Nice to have

  • PhD in AI related research fields
  • developing and deploying AI systems in a real-world setting

What the JD emphasized

  • end-to-end software generation capabilities
  • world’s first end-to-end AI software development agent
  • LLM training and deployment for the software engineering
  • large-scale distributed training
  • model quantization, pruning, or CUDA-based deployment optimization

Other signals

  • LLM training and deployment for software engineering
  • end-to-end AI software development agent
  • model training, optimization, quantization, and deployment