Role Responsibilities：

Analyze real-world business scenarios, identify key challenges in complex document processing, and design deployable algorithmic solutions.
Leverage existing large language and multimodal models to perform targeted optimization, architectural improvements, and performance validation.
Build evaluation, benchmarking, and error analysis frameworks to continuously improve recognition stability and generalization.
Contribute to RAG (Retrieval-Augmented Generation) and agent architecture design, optimizing document understanding and automated decision-making workflows.
Drive models from experimentation to production deployment, addressing performance, latency, and cost optimization.
Collaborate with engineering teams to productionize algorithms and integrate capabilities into enterprise-grade systems.

Requirements：

Master’s degree or above in Computer Science, Artificial Intelligence, or a related field.
Solid foundation in machine learning and deep learning; familiarity with NLP and/or multimodal model applications.
Proven experience delivering algorithms into production, beyond academic or lab-based research.
Proficiency with deep learning frameworks such as PyTorch.
Familiar with model tuning, inference optimization, and production performance analysis.
Ability to break down business problems and define practical paths for algorithmic improvement.
Strong engineering mindset and effective collaboration skills.

Nice to Have：

Experience with enterprise-grade document understanding or information extraction projects.
Familiarity with vector databases (e.g., Qdrant) and hands-on experience with RAG architectures.
Experience improving model stability and/or optimizing inference cost.
Knowledge of model explainability and risk control methods.
Experience building systems for complex B2B enterprise use cases.

Application Process

If you share our vision and have relevant experience
please send your resume to: careers@wentong.cloud
Email subject line: Position Title + Full Name