Active quantitative research for Self-operated Token fund, built hotspot monitoring and token selection model based on support vector machine, LSTM neural network, Temporal attention mechanism, ran CTA trend trading strategy, and achieved a monthly return of 43.2% during the period, with a withdrawal of 5.6%.Python model development collaborating with external AI education institution , completed model orders for customers, including MDP(Markov Decision Process) quotation and dispatching model, reinforcement learning for edge computing, CNN, and NLP theme analysis model.

Participated in the establishment of F3Hash NFT platform, , accumulated 1000+ seed users, and sold 5000+ NFT in total during my tenure after the implementation of operation strategy. Strategic research and support for the company's cloud computing power, FIL/usdt pledged financial management, NFTFi and other product lines, and proposed a short-cycle flexible mining product model. Helped the company to complete the development of external strategic cooperation resources with Xunlei, Zoomlion Digital Science, Qiniu Cloud Storage, Singapore Rockx and other companies, and realized the complete separation of domestic storage business and overseas computing business and structure.Assisted the President's office and the Legal Department to complete the investment of RMB 5 million in the Westworld computing platform.

Completed a number of in-depth analysis reports such as "Libra2.0 Interpretation", "ETH 2.0 In-depth Analysis", "DeFi Finance", "Filecoin Mining Ice and Fire", "NFT Industry Nature", "Blockchain subdivision track Heat Quantitative Model". Supported Huobi Group business unit strategy, including ETH2.0, Filecoin lending business, Huobi Heco public chain economic model and DEFI financial products cold start. Became a lecturer of the blockchain talent training course of the Ministry of Industry and Information Technology, China and passed the blockchain finance specialist certification.

Completed the construction of the overall research module of ZBNexus platform , including industry analysis, project rating, market research and evaluation, and Tokenomics LabPublished "Public chain decentralization Quantitative Evaluation Model", "token Design from the perspective of engineering" and other original hot articles, selected by Blockchain media "the Mars Finance" as monthly best blockchain in-depth articles for many times, the accumulated articles readings is more than 10,000. In the same year ZBNexus was selected by a number of medias as annual blockchain excellent think tank.

Focus on intelligent electronics, artificial intelligence, 5G communication and other industries in the TMT sector, and complete 10+ in-depth reports.The annual return rate of the company's account portfolio was 10%-15%. I tracked 10 listed companies for the fund manager, visited over 30 upstream and downstream enterprises, and recommended several stock targets with a return rate of more than 50%.