Kỹ sư Học máy

TP. Ho Chi Minh

Đăng ngày 19/03/2026

Train and fine-tune models for target use cases (ASR/STT, TTS, OCR, LLM, classification, recommendation).
Prepare, clean and label datasets for experimentation and training pipelines.
Build evaluation frameworks and metrics (for example - accuracy, WER, MOS, F1, OCR precision/recall) and run systematic A/B/benchmark tests.
Analyze failures, hallucinations, bias and weak cases and propose improvements to models and data.
Optimize model performance for inference (quantization, LoRA/QLoRA fine-tuning, prompt optimization, decoding strategies).
Collaborate with platform engineers to package trained models for deployment and monitoring in production.
Document experiments, datasets, evaluation results and share learnings with the team.
Work with business teams to understand feedback, define acceptance criteria and measure real-world performance.

KEY REQUIREMENTS:

3+ years of experience in Machine Learning or related role working with deep learning models.
Solid hands-on experience with Python, PyTorch, HuggingFace Transformers.
Practical experience preparing datasets and running training/finetuning workflows.
Understanding of model evaluation methodologies and experience with metrics/benchmarking.
Experience with one or more of: speech recognition, TTS, NLP/LLM, OCR/document understanding.
Ability to analyze errors and propose iterative improvements based on data and evaluation results.
Experience working in agile product delivery teams.
Fluent English - spoken and written.
Analytical thinking, problem solving, structured approach, attention to detail.
Proactive, curious, willing to experiment and learn continuously.

COMPENSATIONS & BENEFITS:

If you are referred for this position by our Employee/Recruitment Collaborator, please apply via this LINK.

If you are an internal candidate, please apply via this LINK

Otherwise, please click the Apply button as below for application

Actions