Pinduoduo Has Entered the Field of Large-scale Models

Recently, Pinduoduo has established a large-scale AI model team consisting of dozens of people, located in Shanghai.

According to the introduction, Pinduoduo‘s large-scale model team will explore the application of AI large-scale models in customer service, dialogue, and other scenarios. It will also expand to its cross-border e-commerce platform TEMU’s intelligent customer service, search, recommendation, and other business scenarios. Currently, the entire process is still in the research and development stage.

In the official recruitment position directory of Pinduoduo, positions in fields such as LLM and NLP algorithms have already appeared. Most of these positions are located in Changning District, Shanghai, with salaries ranging from 30-60K (approximatly $4,200-8,400). Among them, there are also positions offering 50-80K (approximatly $7,000-11,200) and a 16-month salary.

One of the positions is called “Large-scale Model Algorithm”, and it is described as requiring a deep understanding and practical experience in LLM, multimodal, and multilingual domains. It also requires practical application or project experience in model pre-training, performance tuning, and content generation. Familiarity with common deep learning algorithms such as LSTM, Bert, as well as familiarity with TensorFlow, PyTorch, Keras, MxNet are also required.

SEE ALSO: Pinduoduo’s Cross-border E-commerce Platform Temu Charges A Service Fee of 0.5%

According to the introduction, this position is mainly responsible for:

Researching, training, and applying LLM, including but not limited to tasks such as multilingualism, multimodality, model fine-tuning, deployment optimization, etc.

Responsible for data preprocessing, feature engineering, model training, and model evaluation to ensure the efficiency, accuracy, and stability of the models.

Utilize cutting-edge technologies such as deep learning to conduct dialogue/text/image comprehension tasks, and explore the large-scale application of big models in customer service scenarios.

Track and study the latest developments in the industry of large models, always maintaining technological advancement.