| AI Project | link |
|---|---|
| huggingface link | |
-
π 13+ years in backend & cloud-native architecture (Microsoft, OPPO, Didi, etc.)
-
β‘ Experienced in Java | Flink | Kafka | Kubernetes | Go | Python | Rust |
-
π Passionate about Web3, blockchain protocols, DEX, and on-chain data analytics
-
𧩠Recent focus on LLM Agent & RAG system construction (e.g., DeepSeek/Tushare Agent, Multi-Document RAG), Flink stream processing, and Token Smart Score analytics.
| Project | Description |
|---|---|
| AI Financial Data Agent | Developed an intelligent AI Agent powered by DeepSeek LLM to perform natural language querying and analysis on financial data, automatically leveraging the Tushare API and exposed via a Gradio UI. GitHub Link |
| Multi-Document RAG Engine | Implemented a high-performance Retrieval-Augmented Generation (RAG) system using Hugging Face models and FAISS to efficiently process and synthesize answers from multiple unstructured documents, enhancing knowledge retrieval accuracy. GitHub Link |
| Web3 & Blockchain | Led the creation of a BlockSight on-chain profiling platform to unlock valuable user and token insights across multiple chains, driving new business opportunities in DeFi and marketing through a high-performance blockchain explorer and a token smart scoring system. |
| Kubernetes Multi-Cloud Scheduling | Architected a Kubernetes Federation Scheduler to enable seamless, highly-available deployment across multiple cloud providers, ensuring business continuity and reducing vendor lock-in risk. |
| Flink Operator | Developed a cross-cluster Flink Operator to provide centralized management and orchestration for Flink jobs, significantly improving our DevOps efficiency and ensuring the stability of mission-critical data pipelines. |
| Personalized Push System | Built a scalable push platform that fundamentally changed how we engage users, achieving a 200% increase in CTR by leveraging real-time data and advanced feature modeling, directly impacting user retention and content consumption. |
| Real-time Risk Monitoring | Developed a robust real-time risk monitoring system with 92% accuracy, allowing the business to proactively prevent financial and safety losses. The system's automated processes also improved our operational efficiency by 60%. |
| Domain | Tech Stack |
|---|---|
| AI/LLM | LLM Agents, Retrieval-Augmented Generation (RAG), Hugging Face, FAISS |
| Architecture | Cloud-Native Architecture, High-Concurrency, Distributed Systems, Microservices |
| Languages | Java, Python, Go, Rust |
| Big Data | Flink, Spark, Kafka |
| Cloud | Alibaba Cloud, Azure |
-
Email: [email protected]
-
Phone: +86 18511303365