Jiayu Li

Software Engineer

Chicago, Illinois, United States9 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Developed AI Search system for TikTok.
Expert in full-stack development and machine learning.
Strong academic background with high GPA.

Stackforce AI infers this person is a Full-Stack Software Engineer with a focus on AI and Machine Learning in B2C environments.

Contact

Skills

Core Skills

Ai SearchMachine LearningBack-end Web DevelopmentDistributed Systems

Other Skills

Multi-agent SystemsPythonKafkaHiveRetrieval-Augmented Generation (RAG)Model Context Protocol (MCP)Search EngineWeb DevelopmentDevOpsMLOpsFront-End Development

About

👋 Self Introduction Hello! I'm Jiayu (Judy) Li, an aspiring Software Engineer currently pursuing my Master's in Computer Science at the University of Chicago, focusing on full-stack software development and machine learning. I have extensive experience in full-stack software engineering, along with a strong understanding of cutting-edge ML algorithms and infrastructure. 🎯 What I'm Looking For With a strong passion for full-stack development and AI, I am actively seeking SWE and MLE intern positions for summer 2025. 💻 Professional Experience During my internship at Bytedance, I developed both front-end and back-end systems for Jichuang AI, a platform that utilizes AIGC tools to generate ads for businesses on TikTok. 🛠️ Skills - Programming Languages: Python, C/C++, Go, Java, SQL, TypeScript, JavaScript, HTML, CSS, R, x86 Assembly - Web Development: Flask, Spring Boot, REST, MySQL, Redis, RabbitMQ, WebSocket, RPC, Node.js, React.js - Data Engineering: Numpy, Pandas, Matplotlib, Tableau, Apache Spark, Apache Hive, A/B test - Machine Learning: PyTorch, Tensorflow, Transformers, Keras, Scikit-learn, Weight & Biases, Google Vertex AI - DevOps Tools: Git, Linux, Shell, Docker, Kubernetes, AWS, GitLab CI/CD, Jenkins, Jira - Mathematics: Calculus, Linear Algebra, Probability, Numerical Analysis, Foundations of Machine Learning

Experience

Tiktok

Software Engineer

Jun 2025 – Present · 9 mos · San Jose, California, United States · On-site

Built an AI Search system generating ecom LLM decision cards displayed in Top1 search results on TikTok, architected a full pipeline from offline generation to online retrieval.
Designed and created an Agentic Deep Research workflow based on LangChain & LangGraph in Python; applied the ReAct paradigm with RAG-style retrieval and dynamic tool invocation (e.g. product retrieval via Model Context Protocol (MCP)); parallelized researcher agents to cut overall latency by ~40%; and established end-to-end traceability by streaming all workflow traces to Kafka and materializing them into Hive tables, enabling efficient case-level debugging and historical analysis.

AI SearchMulti-agent SystemsMachine Learning