James(Meng) Lee AI Agent Engineer at SoluLab

Senior AI/ML Engineer

SoluLab

Jan 2023 - Current (3 years 6 months)

LLM & Generative AI: Built and deployed RAG systems using LangChain and LangGraph for customer support applications, integrating vector databases (Pinecone, Qdrant) for semantic search. Worked on improving answer relevance and reducing hallucinations through hybrid retrieval approaches. Fine-tuned open-source LLMs (Llama 2 Mistral) using LoRA/QLoRA techniques for cost optimization. Experimented with model quantization (GGUF, int8) to improve inference speed. Worked on optimizing vector search performance using FAISS and Qdrant, reducing retrieval latency for document search use cases. Built automated review summarization system using fine-tuned transformer models for product sentiment analysis. Worked on multi-agent AI system using LangGrap

Machine Learning & Data Engineer

Andersen Lab

Aug 2020 - Dec 2022 (2 years 5 months)

Data Infrastructure & ETL Pipelines: Built and maintained ETL pipelines using Apache Kafka, AWS Kinesis, and Spark for processing large volumes of event data. Set up data lake infrastructure on AWS S3 with Redshift for analytics, working on data partitioning and query optimization. Built batch and real-time data processing pipelines for multiple client projects. Worked on migrating legacy systems to cloud-native architecture on AWS and Kubernetes.Feature Engineering & Data Quality: Developed feature engineering pipelines for fraud detection and other ML use cases. Built data quality checks and validation pipelines for production ML models. Worked on feature versioning and lineage tracking for ML workflows.ML Production & Operations: Product

A

Software Engineer

Archr

Jul 2019 - Jul 2020 (1 year 1 month)

Developed full-stack features using React.js, TypeScript, and Python (FastAPI), building responsive UIs and RESTful APIs for client integrations. Created data visualization dashboards using React.js and D3.js, integrating with PostgreSQL backend. Developed ML classification models (Random Forest, SVM) for ESG scoring and built data processing pipelines using Pandas and SciPy. Integrated ML models into web application, creating API endpoints for real-time scoring. Contributed to microservices architecture using Docker and Kubernetes, assisting with CI/CD pipelines and AWS services configuration.

U

Full Stack Engineer

Unit 4

Oct 2017 - Jun 2019 (1 year 9 months)

Developed full-stack ERP modules using React, Django, and PostgreSQL, building features for HR, finance, and payroll operations. Built RESTful and GraphQL APIs for integrations with external systems, implementing endpoints under senior engineer guidance. Worked on frontend state management using Redux and implemented user authentication features with OAuth and SAML. Developed Django backend services, implementing business logic, API endpoints, and background tasks using Celery. Participated in migration to Azure cloud infrastructure with Kubernetes, assisting with CI/CD pipelines and Docker containerization. Integrated AI-powered microservices for expense classification and worked on API integrations with ML services.

James(Meng) Lee

About

Detail

Timeline

Résumé