Senior Product Manager - AI-Driven Internal Platform

Gritty diffusion and adoption. You are the internal evangelist and process engineer rolled into one. You will sit side-by-side with ops teams, map their workflows, write the English-language SOPs, inject engrams, and manually bridge the gap between their daily chaos and our automated platform.

Define the workflow primitive library. You will look at messy corporate processes and instinctively break them down into a small set of reusable workflow types what we call the workflow alphabet. You will enforce our core rule: the process wins, and the platform exists to serve the process.

Own the feedback loop. Because you are building the initial workflows yourself, you will personally feel the friction. You will translate those pain points into immediate feature enhancements for the core engineering team.

Opinionated telemetry and metrics. You will define how we measure success. You will think deeply about token budgeting, context window efficiency, system drift, and error/supervisor rates, establishing the telemetry needed to prove operational ROI.

Hold the line on system design. You will protect the platform's core invariants  stateless execution, context-constrained prompts, engram-driven knowledge  even when shortcuts are tempting. The discipline that makes this platform work compounds; the shortcuts that erode it compound too.

You use AI extensively in your own work. Prompting, drafting, prototyping. You can ship a working n8n prototype in an afternoon and you treat AI tools as part of your default workflow, not as a thing you study from a distance.

First-principles AI understanding. We don’t care if you have "3 years of official AI PM experience"  that's usually fluff. We care that you deeply understand how LLMs actually work. You must grasp context windows, token budgeting, deterministic vs. probabilistic execution, and why grounding a stateless model prevents hallucination.

Systems and operations mindset. You possess an instinctive love for systems thinking. You can take a disorganized, human-centric process and cleanly map it into structured, modular data.

Extreme execution and grit. You don't view workflow configuration, writing SOP documentation, or manual testing as "beneath" a product leader. You want to get your hands dirty to understand the core mechanics of the system.

Technical depth. You can speak the language of system architecture, API design, data schemas, and state management just as fluently as you speak to a non-technical operational specialist.

Deep comfort in long-form writing. Coda operates in writing. Your first months will involve writing more than most PMs write in a year — PRDs, SOPs, engram drafts, decision logs, internal explainers. If that sounds tedious, this is the wrong role.