Bjakcareer
Technical Lead, Machine Learning
China, Remote. Posted 4 days ago
The role
This person will design and build ML systems covering data processing, model training, evaluation, inference and deployment. They will develop training and fine-tuning pipelines for large language models and guide engineers through technical design and code reviews.
What you will do
- Design and build ML systems covering data processing, model training, evaluation, inference and deployment
- Develop training and fine-tuning pipelines for large language models
- Build evaluation frameworks to measure model quality, reliability and performance
- Optimise inference systems for latency, throughput, GPU utilisation and cost
- Develop data pipelines for collecting, processing and validating training and evaluation datasets
- Deploy, monitor and troubleshoot ML models and infrastructure in production
- Make architecture decisions and guide engineers through technical design and code reviews
- Work with research and application engineering teams to improve model performance in production
Required qualifications
- 8+ years of software engineering or machine learning engineering experience, including production ML systems
- Strong Python programming skills and hands-on experience with PyTorch or JAX
- Experience training, fine-tuning and evaluating large machine learning models
- Experience building and operating GPU-based training or inference systems
- Strong understanding of distributed computing, data pipelines and ML infrastructure
- Experience optimising model performance, inference latency and resource utilisation
- Ability to design ML system architecture, review code and resolve complex production issues
- Experience providing technical leadership to ML engineers
Nice to have
- Experience with large language models, transformers and distributed model training
- Experience with inference frameworks such as vLLM or TensorRT-LLM
- Experience with synthetic data generation, data curation and model evaluation pipelines
- Experience deploying ML systems for consumer applications
About the company
bjakcareer aims to build proactive applications for anyone in the world with minimal to no prompting. Their first application is an agent-first mobile communication app.