I design and build production-grade large-model agent systems and multimodal AI pipelines. My work focuses on translating cutting-edge LLM research into scalable engineering systems, with an strong emphasis on agent orchestration, multi-agent collaboration, and real-world deployment.

I specialize in architecting end-to-end intelligent agent ecosystems — including memory systems, tool integration, RAG pipelines, multimodal processing, and infrastructure optimization.

AnyGen is a multi-agent AI system designed for complex generation workflows and collaborative reasoning.

Key contributions:

• Designed a multi-agent architecture enabling role-based agent collaboration
• Built tool-augmented LLM pipelines for task decomposition and execution
• Integrated multimodal capabilities (text / image workflows)
• Implemented scalable backend infrastructure for real-time interaction
• Optimized agent memory and retrieval mechanisms for long-context tasks

Tech highlights: Python · LLM orchestration · Agent frameworks · Multimodal pipelines · Docker · Distributed backend

Focus: bridging research-grade agent design with production-ready engineering

• Large Language Model (LLM) agent architecture & orchestration
• Multi-agent systems and collaborative workflows
• Multimodal AI pipelines (text / image / speech)
• Retrieval-Augmented Generation (RAG) systems
• Agent memory and reasoning optimization
• AI infrastructure and deployment engineering

Languages Python · Bash

LLM & Agent Systems Agent orchestration · Prompt engineering · RAG pipelines · Tool integration · Multi-agent coordination

Machine Learning & Multimodal PyTorch · Transformers · Multimodal model pipelines

Infrastructure & Engineering Docker · Backend system design · Scalable AI services · API architecture

• Advanced multi-agent collaboration frameworks
• Production-ready multimodal agent pipelines
• Performance optimization for large-scale LLM systems
• Intelligent workflow automation with agents

• Long-term agent memory architectures
• Autonomous agent planning and reasoning
• Scalable LLM deployment strategies
• Emerging multimodal foundation models

• GitHub: https://github.com/zhanglina94
• LinkedIn: (https://img.shields.io/badge/-LinkedIn-blue?style=flat&logo=Linkedin&logoColor=white)](https://www.linkedin.com/in/lina-zhang-58440b101/)
• Email: (https://img.shields.io/badge/-Gmail-c14438?style=flat&logo=Gmail&logoColor=white)](zhanglina249@gmail.com)

⭐ I’m interested in collaborating on advanced agent systems, multimodal AI, and real-world LLM applications.