Jacob Dietle is a context engineer and the founder of 📝Taste Systems, a solo consultancy for venture-backed companies. He studied information systems in college, worked at a few startups, and founded his own agency, which evolved from custom dataset build-outs into 📝go-to-market engineering — automations and AI workflows for clients. The more he used AI on that work, the clearer it became that the bottleneck on AI's capabilities was its context, and that realization became the focus of his consultancy, where he now helps companies build 📝Context Engineering engines to maximize AI's impact across high-value GTM use cases.
Jacob argues context is what lets AI produce specific, relevant output instead of falling to the lowest common denominator. He built a system that cut a client's newsletter creation from roughly six hours to thirty minutes by pulling from high-signal sources into a repo that Claude Code uses as context, and he keeps a human in the loop to validate output, seeing AI as an amplifier of an expert rather than a replacement. On build versus buy for a context OS, he recommends both — tools like 📝Octave give a GTM context graph out of the box, while building your own affords customization and deeper understanding. He advocates managing context the way an engineering team ships code: versioned, reviewable, and collaborative.
Jacob came on the podcast for a context engineering deep dive — how context shapes AI output, how to build a context engine's foundation, and why he manages context the way engineers ship code.