This is Jeffrey Lin.

Finance needed to see what each department was spending on Snowflake before they could set budgets and allocate costs against them. I built an AI agent that answers that in plain language, showing each person only what their role allows. The same interface reaches the HR roster, internal sites and knowledge base, so new starters stop hunting for answers, and I designed the end-to-end pipelines in Snowflake and SQL underneath it.

Less building on this one, more watching how it actually works. I spent it inside a company already shipping data and AI projects, seeing what happens to that work when it meets a real business, and sitting down one-on-one with senior leaders to ask how they choose what matters and get teams across the company pulling the same way.

My job was to find the story inside large datasets, the trends and the anomalies that turn out to matter, then hand shareholders something they could act on rather than a spreadsheet they would put down. I also dug into AI-focused venture capital and wrote up what I found, supporting the firm’s fund-of-funds strategy.