Associate-Tax Technology AI Workspace - Tax - Chengdu

Description The AI Development Assistant is responsible for supporting the design and implementation of AI solutions for knowledge management and intelligent consulting systems. This role focuses on building and optimizing specialized knowledge bases, developing question-answering agents using frameworks such as LangChain and Dify, and ensuring high-quality data processing for documentation purposes. The assistant collaborates with experts and engineering teams to deliver compliant, efficient, and user-friendly AI tools, enhancing workflow automation and decision support capabilities. Expectations: Model Development & Delivery • Lead the design, implementation, and deployment of LLM-powered applications, from conceptualization to production. • Build and maintain frameworks for prompt engineering, model evaluation, and workflow orchestration (e.g., LangGraph, LangChain). • Ensure rigorous testing, validation, and quality control of generative AI models, using tools such as EvalScope. Observability & Monitoring • Integrate LLM Observability for monitoring, troubleshooting, and optimizing LLM application performance, cost, and reliability. • Instrument applications for end-to-end tracing, token usage measurement, latency analysis, error detection, and operational metrics. • Build dashboards and alerts for model drift, performance anomalies, and generative AI metrics. Research & Continuous Improvement • Maintain updated knowledge of generative AI techniques, frameworks, and open-source libraries (e.g., Hugging Face, GitHub, PyTorch). • Lead and participate in research activities to advance LLM capabilities and industry best practices. Collaboration & Stakeholder Engagement • Work closely with cross-functional teams to integrate LLM technologies into products and services. • Engage business stakeholders to identify opportunities for innovation and value creation. Engineering Excellence • Drive adoption of Agile, DevOps, and CI/CD practices for reliable, scalable delivery. • Advocate for automated testing, infrastructure-as-code, and continuous monitoring. • Conduct code reviews and enforce secure coding practices. Skills Set: Technical Skills • Strong hands-on experience in machine learning and generative AI, with a track record of delivering complex solutions to production. • Deep expertise in LLM frameworks and orchestration tools: LangChain, LangGraph, EvalScope. • Proficiency in Python software development, following object-oriented design patterns and best practices. • Experience with deep learning frameworks: TensorFlow, PyTorch. • Familiarity with Hugging Face Transformers, spaCy, Pandas, and other open-source libraries. • Experience with Docker, Kubernetes, and cloud platforms (Azure, Alibaba Cloud). • Knowledge of Postgres, vector databases, and scalable data architectures. • Strong understanding of MLOps, CI/CD, and infrastructure-as-code. Observability & Monitoring • Hands-on experience in instrumenting LLM applications for observability using Langflow and related tools. • Ability to build and maintain dashboards, alerts, and metrics for model performance and reliability. Collaboration & Communication • Excellent problem-solving skills and creative thinking. • Strong communication and collaboration abilities, with experience working effectively in cross-functional teams.