Using DuckDB to Optimize Snowflake and the New Guard of Data Startups (ft. Kyle Cheung)

If you're a data tech startup don't do your DevRel alone. Reach out to Early Signal for bespoke DevRel content and distribution strategy for your GTM. www.earlysignal.tech In this episode, I sit down with Kyle Cheung, co-founder of Greybeam, to discuss his journey from data analyst to data startup founder. Kyle shares how he broke into data analytics with a creative video application, his experience building data infrastructure, and why Greybeam pivoted from an AI data analyst to solving Snowflake cost optimization with DuckDB. We dive into: - The challenge of marketing technical products to both technical and business audiences - Greybeam's memorable toilet analogy marketing campaign and why taking creative risks matters - The pivot from AI data analyst to query routing infrastructure - Balancing personality-driven marketing with enterprise credibility - Why LLM accuracy wasn't ready for AI analysts in early days Timestamps: 0:00 - Introduction 1:00 - Kyle's background and journey to startups 3:18 - The creative video application that landed his first BI role ]6:22 - Transitioning from BI/finance to technical founder 9:35 - The balance between technical depth and marketing 11:27 - The toilet analogy marketing campaign 12:24 - Taking creative risks in data marketing 13:45 - Balancing personality with enterprise credibility 17:15 - Why early LLMs weren't accurate enough for AI analysts