Chen Zhizhao

Developed RFQ-Agent multi-agent AI system for cross-border supply chain from scratch, bridging US brand enterprises and Chinese manufacturers

• Achieved universal product category coverage across all merchandise types
• Designed interactive multi-turn dialogue agent averaging 5 rounds to comprehensively clarify buyer requirements
• Innovated multi-role agent architecture leveraging context engineering techniques to replace conventional product attribute knowledge graphs
• Collaborated with engineering team on end-to-end data infrastructure and agent implementation

Led a 5-person algorithm team for AlimeBot (intelligent customer service product), driving end-to-end development from system architecture to model implementation

• Developed UniIntent engine (2023) - industry-first unified intent understanding system based on large-scale pre-trained language models, achieving 6-12 month technical lead over competitors; improved customer satisfaction by +2pt, resolution rate by +0.5pt, with 1000 daily QPS at 30ms CPU inference latency
• Built retail-domain dialogue LLM (2023) based on Qwen-13B with billion-scale domain data, covering multi-turn understanding, product Q&A, and API integration; achieved 93.5% end-to-end accuracy in product Q&A and 88% in long-tail queries, reducing merchant operational costs significantly
• Pioneered retrieval-augmented generation approach for e-commerce dialogue systems (2022-2023), implementing knowledge-enhanced LLM to mitigate hallucinations across pre-sales, post-sales, and recommendation scenarios
• Delivered 30+ patents; established comprehensive dialogue algorithm infrastructure including intent classification (2000+ intents), product Q&A, multi-turn dialogue, and knowledge mining platforms

• Led 3-person team managing AliMe's foundational dialogue and predictive recommendation platform, serving dozens of Alibaba Group BUs with 10M+ daily user interactions

• Launched streaming-based proactive service platform (2019) for real-time buyer question prediction, significantly improving customer satisfaction across all touchpoints

• Pioneered recommendation-driven dialogue experience (2018) by introducing deep CTR and GNN models for "Guess What You Want to Ask" feature, and explored RL-based interactive recommendation strategies

Joined as early founding team member, leading the development of DtWave's core data platform and algorithm-related infrastructure

Led consumer intelligence initiatives including TCIF buyer profiling system (100+ algorithmic tags), industry-first annual review product using topic modeling, and personalized product recommendation for Taobao Search (price prediction & category matching)