Skip to content
View YingzuoLiu's full-sized avatar
3 followers · 12 following

Block or report YingzuoLiu

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
YingzuoLiu/README.md

Yingzuo Liu

AI Agent / LLM Runtime Engineer focused on building reliable multi-step AI systems.

I work on:

  • Agent runtime architecture: Planner / Executor / Validator
  • Typed AgentState, StatePatch, reducer-based state updates
  • Tool calling, workflow validation, retry, blocker propagation
  • RAG pipelines, retrieval evaluation, hallucination control
  • Python, FastAPI, PyTorch, HuggingFace, FAISS, Redis, Docker

Featured Projects

1. Travel Agent Runtime Demo

A runnable multi-turn travel planning Agent runtime focused on state consistency, partial replanning, deterministic validation, and trace logging.

Key ideas:

  • Typed AgentState with Pydantic
  • StatePatch + reducer
  • Partial replanning
  • Validator-based hard constraint checking
  • FastAPI service entrypoint
  • Traceable execution

2. ReAct Quant Research Agent

A ReAct-style quantitative research agent with workflow planning, market data retrieval, factor computation, ranking, pipeline validation, and repair loop handling.

Key ideas:

  • Planner-visible step catalog
  • Workflow construction
  • Runtime execution
  • Required-step validation
  • Idle detection and repair prompts

3. RAG / Retrieval Systems

Projects exploring document ingestion, embedding retrieval, reranking, answer grounding, and evaluation metrics such as Hit@K, Recall@K, MRR, and hallucination checks.

Current Focus

I am especially interested in:

  • Agent runtime / harness engineering
  • Long-horizon task control
  • Skill-based tool orchestration
  • Evaluation-driven Agent systems

Pinned Loading

  1. travel-agent-runtime-demo travel-agent-runtime-demo Public

    Python

  2. RAG-Pipeline-with-LlamaIndex RAG-Pipeline-with-LlamaIndex Public

    Retrieval Augmented Generation Pipeline with LlamaIndex and DeepSeek

    Python

  3. chatbot-for-cross-border-e-commerce chatbot-for-cross-border-e-commerce Public

    FastAPI + A2A + MCP stub + RAG-TFIDF

    Python

  4. multimodal-search multimodal-search Public

    A production-ready multimodal search system for e-commerce

    Python

  5. react-quant-research-agent react-quant-research-agent Public

    Python

  6. rec-search-engineering rec-search-engineering Public

    Python