Consulting

I take on a small number of short-term, intense “missions” each year. The goal is simple: take one gnarly problem and turn it into a shipped solution (or a crystal-clear path to one), with a real definition of done.

Over the years, I’ve realized my startup success is less about “business skills” and more about how I solve problems. I’m a generalist / technologist with deep roots in science and engineering, and I use first-principles thinking, high-bandwidth synthesis, and a bias toward building to turn messy ambiguity into practical outcomes.

Most of my work is AI/product strategy and tool-building for companies. A couple times a year, I also take on applied science/engineering missions with R&D teams and technical organizations. 

What I mean by consulting

Disclaimer: I’m not a traditional consultant, and I don’t do retainer-style advisory. My work is project-based and outcome-driven.

Most missions are fixed-fee or milestone-based. When the outcome is measurable, I’m also open to success-based structures (because I like being paid for results, not hours).

I only take 3–5 missions per year, and I’m picky about fit. If I’m not the right person for the job, I’ll tell you quickly and try to point you to a better option. 

Who I work with

High-agency leaders and small teams on 0→1 or inflection-point problems. I do best in the messy moments where direction is unclear, stakes are real, and you need someone who can think and build.

On the business-side, clients are typically:

• Startups, mid-size companies, and PE-backed operators

• Leadership teams who want AI leverage without “strategy theater”

On the “science/engineering” side:

• Corporate R&D groups, labs, and research orgs

• CROs and technically specialized teams

• Government agencies / programs 
  

What a mission looks like

A mission is tightly scoped and built to avoid consulting creep:

• Scope + definition of done (so we don’t wander into slideware)

• Discovery sprint (constraints, data, “what would break?”)

• Design + build sprint (research, solution exploration, prototyping → working systems)

• Handoff (docs, roadmap, implementation plans, etc)

I usually work solo or with 1–2 trusted collaborators.

What you get

Depending on the mission, the output might be:

• A working internal agentic AI tool/system (not just slides)

• A well-tested solution hypothesis with an evaluation plan

• A clear implementation roadmap (what to do next, in what order, and why)

• A short enablement session for your team
  

When I’m a good fit

• You have a real 0→1 or inflection-point problem

• You want someone who can synthesize quickly and ship

• You have a capable team and can move with urgency

• You’re willing to give real design authority for the duration of the mission
  

When I’m not a fit

• Ongoing retainers or “jump on calls as needed”

• Staff augmentation

• Long-term maintenance of tools I built

• Projects where nothing ships

Examples of Prior Engagements

AI + Business / Product Strategy Missions 

• Designed and shipped a custom ‘AI Investment Copilot’ that ingests a family office’s data sources (portfolio data, memos, financial reports, BI tools, etc.) and outputs a single monthly ‘what changed and what matters’ briefing.

• Developed an AI-first product and business model strategy for an established mid-size professional services company worried about a new AI entrant (what to build, why it wins, how to ship).

• Designed and built a Daily Executive Briefing system that ingests a company’s internal data sources (systems of record, reports, KPIs, etc.), relevant external sources (industry reports, news, etc.), and executives’ personal sources (email, Slack, etc.) and formats them into an actionable Morning Briefing.

• Designed and built an AI Investment Analyst that applies a firm’s specific value-investing framework to evaluate publicly-traded companies — sourcing and synthesizing inputs like 10-Ks, earnings call transcripts, analyst reports, industry research, internal theses, and relevant current events — in order to deliver buy/hold/sell recommendations at or above the quality of the firm’s human analysts.

If you’re looking for a label: this often looks like “fractional Chief AI Officer / CTO / strategy lead” for a defined window of time.

Applied Science / Engineering Missions 

• Delivered a physics-grounded optimal-control framework for fast-charging Li-ion batteries, turning thermodynamic efficiency and degradation risk into an implementable current/voltage waveform optimization problem based on a novel theoretical consideration.

• Designed an experimental plan that turns a research concept into an implementable model that can be practically tested under real-world constraints.

These are usually problems where the team wants an outside-the-box solution that requires a mix of deep research & synthesis, rigorous thinking, and practical implementation grounded in hard science and engineering constraints.

Curious why a generalist can be useful here? I’ve written a couple of posts about the AI-enabled generalist and how LLMs are changing scientific discovery and engineering: Glenn’s Blog

Questions?

Tell me more about the problem you’re trying to solve, why it’s hard, what you’ve tried, and any constraints (timeline, data, systems, stakeholders). I’ll always follow-up and tell you whether I can help or not.