Junxuan Li (William)
AI / ML Engineer(Full-Stack, Backend-focused) • Multimodal RAG • On-device Inference

Junxuan Li (William)

End-to-end systems: multimodal retrieval, agent workflows, evaluation-driven iteration, and mobile-friendly deployment.

Aberdeen, Scotland, UK
Right to work in the UK
Junxuan Li (William)
AI/ML • Agents • RAG • Edge
Focus
Reliable AI systems in production
Strength
Deployment-aware optimization
Domain
Healthcare agents & tooling
Stack
Java • Python • Kotlin
Multimodal RAG + recall evaluation
Agent orchestration (LangGraph / ReAct)
On-device inference (quantization / conversion)

Projects

Selected work with clear scope, metrics, and deployment constraints.

Experience

Production ownership, stability, and performance work across high-concurrency financial systems.

Skills

Engineering-first, with ML systems and tooling as the core.

Education

Recent academic focus on evaluation, agents, and applied AI.

University of Aberdeen — MSc Artificial Intelligence
09/2024–09/2025 • GPA: 16.39/22 (With Commendation) • Class Representative
Symbolic AI Machine Learning Evaluation of AI Systems Applied AI NLG Knowledge Representation & Reasoning Software Agents & MAS Data Mining with Deep Learning

Contact

Open to AI/ML engineering, agent/RAG systems, LLM/VLM evaluation, and on-device inference roles.