Zac Choi is betting that 'dirty data' is the AI industry's biggest hurdle

Choi, a veteran of data infrastructure with two decades of experience, describes the current AI landscape as a mismatch between probabilistic tools and deterministic data requirements. According to the entrepreneur, most corporate data estates were built for human users asking predictable questions, leaving AI agents to guess at the meaning of messy, disorganized inputs. Big Context & Company positions itself as a solution to this problem, essentially acting as a digital sanitation service that cleans and organizes data to prepare it for reliable retrieval.

Building on a career that includes data transformation roles at McKinsey and the successful exit of his first startup, String AI, Choi emphasizes that the primary challenge for modern enterprises is no longer just building software, but mastering data distribution. He warns that first-time founders often focus too heavily on the product while neglecting the market reality. His approach to business dilemmas relies on a framework he calls 'Zero-One-Two-Three,' which prioritizes first-principles thinking and the necessity of human advisors to prevent AI-driven echo chambers. By treating the customer as a co-founder and focusing on scalable, AI-native services, Choi aims to bridge the gap between expensive AI projects and their real-world utility.