T
Timothy Lee
About
Detail
Andhra Pradesh, India
Timeline
work
Job
school
Education
Résumé
Jobs
verified_user
0% verified
Machine Learning EngineerRobloxJun 2025 - Current (10 months)Built a LangGraph-based multi-agent system with real-time discovery, search, ranking, and personalized itineraries on top of scalable AWS microservices and session management.Developed a RAG-based knowledge platform using hybrid retrieval (BM25, TF-IDF, embeddings) and fine-tuned embedding models to serve highly relevant answers with low latency via caching, aligning with large-scale search infrastructure for fast, high-quality results across a massive content catalog.Created a domain-specific assistant that performs real-time legal search and case-precedent ranking using a LLaMA-3 RAG pipeline over a privacy-preserving corpus, showcasing deep expertise in query understanding, relevance modeling, and domain-aware result ordering.Implemented- Machine Learning EngineerAppleNov 2020 - Jun 2025 (4 years 8 months)Engineered a synthetic data generation pipeline to train and fine-tune conversational models on dynamic, customer-driven queries, creating a more chatty, human-like LLM experience and deploying it on SageMaker for real-time interactions with reduced latency.Enhanced a multimodal real-estate chatbot with specialized agents for market-trend retrieval, text-to-SQL, graph visualizations, and contract generation using LangChain, RAG, BERT, LLaMA 3.1, Whisper, and advanced prompt engineering to answer user-specific property questions in text and audio.Built an interactive clothing-visualization system that lets users customize apparel (colors, textures, styles) on themselves or models, using multi-agent workflows, VLM feedback loops, ControlNet,
- Machine Learning EngineerWorkdayMay 2019 - Nov 2020 (1 year 7 months)Designed and constructed a resume-screening platform employing spaCy Named Entity Recognition (NER) to streamline parsing of candidate resumes and job descriptions.Developed a robust question-answering system using NLP and vector database technologies to deliver precise and informative responses over enterprise documents.Built a text-processing pipeline with NLTK and LangChain to extract and analyze data from PDFs and URLs, enabling richer document understanding and retrieval.Utilized the FAISS vector database to store and retrieve embeddings for efficient similarity search, leveraging tools such as SerpAPIWrapper and RecursiveCharacterTextSplitter to enhance retrieval quality and throughput.Developed an asynchronous PDF processing pipeline
Education
verified_user
0% verified
- Master of Science in Computational Science and EngineeringHarvard UniversityJan 2017 - Jan 2019 (2 years 1 month)Cambridge, MA
- Bachelor of Science in Computer ScienceEmory UniversityJan 2013 - Jan 2017 (4 years 1 month)Atlanta, GA