About
About Benjamin
I am a Principal Product Manager who builds and scales products across very different operating models, from lean startup constraints to large, matrixed, regulated enterprises. I use AI and LLMs heavily in my own professional workflow (research synthesis, scenario pressure-testing, spec and acceptance-criteria drafting, and rapid prototyping), always anchored on product judgment, explicit evaluation, and a credible path to production (permissions, data boundaries, human review, and rollback). I do my best work when the environment is messy, the stakes are high, and better judgment earlier can materially change the outcome.
Product philosophy
- Clarity is a product: if stakeholders cannot explain the bet, the sequencing, and the tradeoffs, execution cost goes up fast.
- Decisions should carry evidence, not vibes: research, data, operational context, and explicit alternatives, not generic “vision.”
- Systems beat heroics: platforms, workflows, and integrations need owners, contracts, and measurable health, not one-off pushes.
- AI is a lever, not a substitute for judgment: prompts and workflows should be versioned, testable, and owned like any other interface, especially when operators and revenue depend on the outcome.
What I’m best brought in to do
Situations where range, judgment, and execution discipline compound.
- Define 0-1 platforms and operating models where the path is still forming
- Bring structure to messy product environments without freezing agility
- Stand up AI-augmented workflows end-to-end: crisp intents, prompt patterns, eval hooks, pilots, and production hardening
- Align technical and non-technical stakeholders around constraints and tradeoffs
- Modernize workflows and systems with adoption and reliability in mind
- Identify the next best move under uncertainty with explicit assumptions
- Turn complexity into execution tied to measurable business and operational results
Startup, scale, and enterprise
How different operating contexts shaped judgment, not slogans.
Startup experience taught me speed, pragmatism, and how to ship with incomplete information. Scaling environments taught me alignment, sequencing, and how to make process serve outcomes, not the reverse. Enterprise and regulated contexts taught me stakeholder density, vendor reality, and how to keep delivery credible when non-negotiables multiply. Across all three, LLMs are most valuable when they shorten the loop between hypothesis and evidence, and least valuable when they obscure who is accountable for quality, safety, and adoption. Together, that range sharpens pattern recognition: similar failure modes show up in different costumes.