Resume
James Ding
AI Engineer
Work Experience
- Conducted EDA on 12 years of flight operations data (78K+ flights), identified demand drivers and a COVID-era regime shift, then built and deployed a Prophet forecasting model achieving 2% monthly MAPE in production — targeting 1%+ reduction in $49M annual positioning costs
- Built automated data ingestion pipelines (FL3XX, Salesforce, Foreflight → AWS Lambda → S3 → Snowflake) with end-to-end data quality monitoring and alerting
- Designed a data-driven RBAC workaround for FL3XX's booking platform by building a cleansing and segmentation pipeline with custom business logic mapping user roles (owner, individual, full access) and a manual override layer to manage edge cases
- Built QuickSight dashboards enabling real-time monitoring of operational KPIs, flight demand trends, and fleet utilization for executive and operational stakeholders
- Designed and built Jetty, a multi-agent RAG AI assistant for 400+ employees using AWS Lambda, Bedrock, and DynamoDB — enabling natural-language Q&A over internal documentation, policies, and operational data with tool-use and agent orchestration
- Architected an end-to-end vendor contract management pipeline for Legal — automated document intake, processing, and structured storage via custom Salesforce objects, data quality validation, and governed data models replacing fully manual workflows
Education
University of Calgary
Calgary, AB
University of Illinois at Urbana Champaign
Champaign, IL
Technical Skills
Languages: Python, SQL, JavaScript, Go, R
Machine Learning: XGBoost, Random Forest, scikit-learn, time-series forecasting, classification, feature engineering
Deep Learning: PyTorch, TensorFlow, Hugging Face, CNNs, LSTMs, Reinforcement Learning (PPO)
Data Engineering: Data modeling, dimensional modeling, ETL/ELT pipelines, data quality validation, data governance, Snowflake, DuckDB
Platforms: AWS (Lambda, S3, Bedrock, SageMaker, QuickSight, DynamoDB, API Gateway, CDK), Snowflake, Salesforce, Power BI, Git, GitHub Actions, Docker
AI/LLMs: AI agent development (RAG, multi-agent orchestration, tool use), LLM application design (prompt engineering, agent pipelines), AWS Bedrock, OpenAI-compatible APIs