Mengchen Wang

Responsibilities: Conduct data development, data mining, and develop algorithm models
• Constructed a data mart with 150 features based on corporate attributes and logistics behaviors using the
data warehouse construction process, supporting multiple business scenarios, and increasing the weekly data
mining volume for potential clients from 10K to 100K, while reducing data development time by 10 hours
• Established 40 business scenario labels for customers profiling and integrated them into sales app, helping
sales team identify the right customers and resulting in monthly customer development from 5K to 20K
• Applied collaborative filtering and content-based recommendation algorithms to recommend business partners
for 2M clients in the Metaverse Exhibition Hall App, resulting in a monthly growth of 5K new users
• Utilized NLP and Classification models including NER, DBSCAN, and KNN to build an Enterprise
Upstream-Downstream Knowledge Graph for over 10M companies using logistics and bidding data to
identify potential clients, resulting in an elevated sign-up conversion rate from 1% to 5%
• Created and packaged functional User Defined Function (UDF) tools using Java, including a UDF for
encrypting and decrypting mobile phone numbers and addresses using an external secret key acquired
through API, facilitating the implementation of data desensitization in the data mining process