Cathy Liu

WORK HISTORY

EDUCATION

ABOUT CATHY LIU

I’m an AI Solutions Architect focused on designing and scaling practical, production-ready AI systems for legal and finance & accounting teams.With a dual background in Computer Science (UBC) and Commerce (UofT), I specialize in bridging advanced AI capabilities with the real-world constraints of regulated, high-stakes domains — where accuracy, transparency, and operational reliability matter as much as innovation.In my current role, I work on end-to-end AI solutions that help legal and financial professionals:• extract insight from large, complex document corpora• reduce manual review and operational risk• adopt LLM-based systems with confidence and clear guardrailsMy work spans LLM system architecture, retrieval-augmented generation (RAG), cloud infrastructure, and product design, with a strong emphasis on building systems that scale sustainably — both technically and organizationally.What I focus on in practice:• Leading domain-specific AI initiatives such as the Contract Census project for large-scale legal document analysis• Designing AI-driven reconciliation systems for AR/AP and Fixed Assets, reducing manual review and improving data reliability for finance teams• Converting legal and financial workflows into auditable, production-ready AI systems with clear human-in-the-loop controls• Scaling and documenting LLM architectures so solutions are reliable, cost-aware, and reusable across teamsI’m particularly strong at pattern recognition and systems thinking — identifying what actually drives accuracy, adoption, and long-term maintainability, and making those insights explicit for engineers, product teams, and stakeholders.Beyond execution, I care deeply about:• responsible AI adoption in regulated domains• clear communication across technical and non-technical teams• building internal frameworks that outlast any single implementationI enjoy working with teams who want to move AI beyond experimentation and into durable, high-impact production systems — especially in legal, finance, and accounting contexts.If you’re thinking about how to design, evaluate, or scale AI solutions in these domains, I’m always open to a thoughtful conversation.