Frequently Asked Questions

Nils Liu (劉岦崱) builds AI Agent platforms that run inside real financial institutions. He currently leads an overseas AI team at a major Taiwanese financial holding, taking products from 0 to 1 with a cross-functional team. His work centers on production systems under real regulatory and compliance constraints, not demos.

AI Agent platform consulting, GenAI architecture, rapid prototyping, and team training for financial institutions. The focus is multi-agent systems, financial compliance AI, and RAG/GraphRAG enterprise deployment.

EnterpriseGPT is the bank's first production GenAI platform, built by Nils and his team. It went live in August 2024 and serves the entire organization.

AI Coach is an intelligent coaching system for financial advisors, built on a 3-layer Agent architecture with GraphRAG retrieval and rubrics-based evaluation. It reached 96% advisor satisfaction and scaled personalized feedback to 100+ financial advisors.

A scalable, compliant AI Agent infrastructure deployed in the Vietnam market. It supports multiple business scenarios through standardized tool contracts and measurable evaluation loops in a new regulatory environment.

Six granted utility patents, all in LLM application system architecture, with Nils as sole inventor. Two additional invention patents are under review. Each one came out of a real enterprise problem rather than an academic exercise.

Multi-agent systems, financial compliance AI, RAG and GraphRAG, LLM application architecture, and enterprise GenAI platforms — the engineering and trust layers that make AI work inside regulated organizations.

Nils is open to advisory work, speaking invitations (7 conference talks so far), and longer-term opportunities, especially around deploying AI Agents in regulated financial environments. Reach him via LinkedIn (linkedin.com/in/nilsliu) or email at wulingteen@gmail.com.