AI/ML
A collective term for artificial intelligence and machine learning, covering areas such as model training, fine-tuning, inference, data science and agent development. It is a core technology field in the current technology industry, with strong demand for related roles.
Timeline (last 90 days)
AI/ML-related positions are still predominantly Python-based, and job descriptions in areas such as data science, model fine-tuning, RAG pipelines, and Agent framework R&D almost all require Python skills by default.
The advice for job seekers is to build one or two projects demonstrating end-to-end capability (from data processing and model training to deployment) to form a portfolio, which is more impressive to AI/ML teams than a Big 4 background.
AI/ML is a highly hands-on discipline that cannot be truly mastered through reading alone; it requires reinforcement through Kaggle competitions, reproducing paper experiments, or coding exercises.
AI/ML positions have solid requirements for mathematical foundations (linear algebra, probability and statistics) and algorithmic principles, and it is difficult to go far by relying solely on calling existing libraries.
All Facts (4)
AI/ML-related positions are still predominantly Python-based, and job descriptions in areas such as data science, model fine-tuning, RAG pipelines, and Agent framework R&D almost all require Python skills by default.
50%UnverifiedThe advice for job seekers is to build one or two projects demonstrating end-to-end capability (from data processing and model training to deployment) to form a portfolio, which is more impressive to AI/ML teams than a Big 4 background.
50%UnverifiedAI/ML is a highly hands-on discipline that cannot be truly mastered through reading alone; it requires reinforcement through Kaggle competitions, reproducing paper experiments, or coding exercises.
50%UnverifiedAI/ML positions have solid requirements for mathematical foundations (linear algebra, probability and statistics) and algorithmic principles, and it is difficult to go far by relying solely on calling existing libraries.
50%