# Polyiota Profile Last reviewed: 2026-08-29 ## Short Description Polyiota helps companies make consequential work and evidence legible, redesign decisions, and build AI systems that improve as the business changes. ## Founder Daniel Hatke is an operator and advisor based in Colorado. He worked at two hedge funds, reaching the role of chief technology officer, and studied finance and investing at Columbia Business School. He operates two e-commerce businesses. Through Polyiota, he helps companies decide where AI belongs, redesign the work around it, and build the first implementation. That work has also helped him improve the systems behind his own businesses. His newest venture is Kabu Research, an AI-native research platform for Japanese small- and mid-cap equities. He writes The Prometheus Dispatch and co-hosts the Unqualified Advice podcast. - The Prometheus Dispatch: https://prometheusdispatch.com - Unqualified Advice podcast: https://unqualifiedadvicepodcast.com ## Related Work - Two ecommerce companies: Daniel operates two ecommerce companies, including Picture Hang Solutions. They keep the work concrete: decisions meet inventory, pricing, customer service, and the daily cost of systems that do not fit the work. - Picture Hang Solutions: Professional picture-hanging hardware for homes, framers, galleries, and commercial teams. - Kabu Research: AI-native research for under-covered Japanese equities, built around source traceability and explicit evidence. https://getkabu.com ## AI transformation advisory Daniel works with leaders and operators to find the evidence behind consequential work, make the decision logic explicit, and build a governed system that improves as the company changes. ### Make the business legible before automating it. - Work: What happens, who participates, where judgment enters, and where work stalls. - Evidence: Where the necessary data and knowledge live, who owns them, and how reliable they are. - Decisions: Which factors influence choices, which rules are fixed, and where conditions or exceptions change the answer. - Outcomes: What proves success, what reveals failure, and how the company currently learns from either. ### Decide. Design. Build. 1. Decide — Trace the work, source data, decision factors, outcomes, and accountable owner before choosing a system. 2. Design — Set what the system can see and do, where judgment remains human, how outcomes are verified, and how assumptions are challenged. 3. Build — Put a bounded system into use, measure the real outcome, and turn verified experience into controlled improvement. ### Improve the system without hardening the past. The first implementation is a starting hypothesis. Outcomes, counterexamples, and operator review determine what changes—and what should be retired. - Observe: Compare the expected outcome with what happened in the work. - Challenge: Surface counterexamples, exceptions, and changes in operating conditions. - Govern: Let a named owner approve, test, or retire each change before it enters the system. ### What the company can use and own. - An evidence and decision map: The work, source systems, operating knowledge, decision factors, conflicts, gaps, and outcomes that need to become legible. - A governed operating design: The human and machine roles, context, interfaces, authority, verification, exception handling, and learning process the system requires. - A working first loop: A bounded implementation connected to real work, with outcome evaluation, operator intervention, and a controlled path for improvement. ## Canonical Site - Website: https://polyiota.com - Thinking: https://polyiota.com/thinking/ - Advisory: https://polyiota.com/advisory/ - About: https://polyiota.com/about/ ## Contact - Email: dh@polyiota.com ## Structured Data Notes - All pages publish Organization JSON-LD. - The Organization founder references https://danielhatke.com/#person. - All Organization data derives from a single source: src/data/site.ts.