Field notes · Direction · July 2026

Where I'm at — a conversation with Grok Build to find direction

Leaving the Navy in October 2027. Nuclear-trained officer. Building with AI. Not looking for another job. I asked Grok Build to look at what I've already shipped and help me set up the next year for higher-leverage entrepreneurship — not another W-2.

2026-07-28 Career / leverage Grok Build conversation

1. What I asked Grok

Stef → Grok Build

Prompt (summarized)

I'm getting out of the Navy in October 2027. It is July 2026. I'm a nuclear-trained officer. I've been building a lot with AI — look around the repos. Quintile is published and open source; Mission Control and Salesforce automations show what I can ship under hostile constraints; billing / eligibility work shows a real buyer path.

I previously ran a crypto mining farm. I want to replace my income after I leave — without getting another job. Entrepreneurship: raise money, create tremendous value, higher leverage, more buy-in than someone else's org chart.

Near-term (next year): put myself in a position where I don't need to worry about where the next meal comes from — while I pursue higher-leverage ideas. Monetizing the automation I've already proven is part of it; focus is the hard part.

What follows is Grok Build's reply, lightly edited for the public web: net-worth figures removed, conclusions kept. First person is me (Stef). Second-person advice is Grok talking to me.

2. Where I actually am (July 2026)

Grok Build

Timeline: ~15 months of O-3 nuclear officer pay until ~October 2027.

Balance sheet in plain language (no dollar line items): liquid investment accounts, a small crypto stack, retirement savings that should stay untouched until a post-separation plan exists, one rental that cash-flows modestly with real equity, and another rental that is roughly a drag because of rate. Navy W-2 is a time-limited subsidy for building.

Runway math (the honest version): You do not need to replace full O-3 total compensation on day one to “not worry about food.” You need a household burn-rate floor and owned revenue that does not vanish when you stop billing hours. Liquid assets are a backup, not the plan. The plan is: keep the Navy paycheck while building cash-flowing product ownership, then separate with product revenue, non-dilutive capital (SBIR), or a raised seed — not another job.
Critical reframe: You are not a broke founder trying to get rich. You are capital-supported, distribution-poor, and idea-scattered. Crypto mining taught ops and capital allocation. The last year of AI work proves you can ship under hostile constraints. What's missing is one customer who pays you every month for something only you can deliver fast.

3. What the repos actually say

You don't have a skill gap. You have a focus gap.

Pattern Evidence Commercial meaning
Automation inside locked systems Mission Control (forms → Excel/VBA/Acrobat under hard constraints); Salesforce call logger (session-in-browser, no cookie sync) “AI that works where IT forbids AI”
Healthcare revenue ops AccessBlue bulk eligibility, clearinghouse research, local HIPAA voice agent, biller deductible UI requirements Real buyer pain, real design-partner path
Product craft + distribution Quintile — signed, notarized, Homebrew, public OSS, demos You can ship consumer-grade software end-to-end
Capital markets tooling Buffett 2.0 / Dhando stack (now archived on this site) Interesting, crowded, weaker founder–market fit than compliance/domain edge
Deep tech / defense Passive sonar UUV SBIR concept (operator + signal processing biography) Biography-irreplaceable; slow; capital-heavy; highest ceiling

Unfair advantage is not “I use AI.” It is: operator in regulated environments + systems brain + absurd shipping velocity with agents.

Working MTP framing: “Make AI work inside the walls where it's not supposed to go.” Compliance-locked environments stay unsolved because domain experts can't build and builders won't take compliance seriously. You sit in that gap.

4. The two-layer model (next 12 months)

Do not pick one path and starve. Run a dual stack:

Layer A — Floor

Cash + proof you can sell. Something that generates meaningful monthly revenue by mid-2027 without becoming a job.

Layer B — Ceiling

One high-leverage bet you prepare to scale after you leave — or via non-dilutive capital, carefully, while still in.

What Layer A is not: endless freelancing as identity.
What Layer A is: paid design partners who fund the product you'll own. Productized revenue from 1–3 customers is rocket fuel — not a career.

5. Ranked options (for this biography, not generic advice)

Tier 1 — Highest expected value

  1. Healthcare eligibility / pre-visit billing product (AccessBlue + clearinghouse + optional voice fallback). Buyer: small medical billing shops and practices. Pain is concrete: “call payer for deductible before visit.” Monetize seats or practice-level monthly. Fastest path from work already in motion.
  2. “Mission Control for regulated workflows” wedge — insurance / federal-adjacent HR / medical credentialing. Same moat shape as Navy automation: forms, CRM logging, locked environments, PII, no admin rights. Do not productize government workflow publicly while in uniform. Sell the pattern to civilian analogs.
  3. Defense / acoustic SBIR (passive sonar UUV concept). Non-dilutive Phase I → Phase II path. Best ceiling where BQQ-class operator experience + MIT signal processing is non-copyable. Not the “replace paycheck in 12 months” path. The “optionality without selling the whole company” path.

Tier 2 — Useful, not primary

Tier 3 — Park or kill

6. “Sufficient income” without ego math

Define three numbers with the household this week — not LinkedIn cosplay:

  1. Post-Navy monthly floor — housing, food, insurance, rental drag, childcare if relevant.
  2. Comfortable monthly — same lifestyle feel without Navy benefits stacking.
  3. Freedom monthly — investable surplus after living.
Goal Shape How by Oct 2027
Survival floor Covers essentials without panic Product MRR + rental net + tiny draw only if needed
Comfort Near O-3 household feel without the uniform Growing product MRR + rentals + buffer
Replace O-3 + more Surplus and options Product scale and/or SBIR Phase II / seed

You do not need full officer total compensation as a W-2 to feel free. You need owned revenue that doesn't disappear when you stop billing hours.

7. 12-month operating plan (Jul 2026 → Jul 2027)

Phase 0 — Stabilize the board (this month)

  1. Write the one-pager financial floor (the three numbers above).
  2. Kill or freeze Buffett-as-company, random demos, feature sprawl.
  3. Pick one floor product: healthcare eligibility / biller tool is the strongest “already in motion” candidate.
  4. Pick one ceiling track: SBIR passive acoustics or compliance-agent platform narrative — not both at full intensity.
  5. Ethics / legal while active duty: outside employment rules, DoD contracting, using government time/systems, fundraising while in uniform. Get a clean answer from an ethics counselor / transition lawyer before you invoice or form a company that looks like federal work. Non-optional for a nuclear officer.

Phase 1 — First dollar (Aug–Oct 2026)

Goal: one paying design partner, not a perfect product.

Parallel (low hours/week): harvest a sanitized Mission Control case study (metrics, architecture, fabricated data only) for narrative and fundraising — not public workflow demos without clearance.

Phase 2 — Productize the pattern (Nov 2026–Mar 2027)

Goal: meaningful MRR or equivalent pilot cash; 2–3 logos.

Phase 3 — Capital optionality (Jan–Jun 2027)

Path When it makes sense
Bootstrap SaaS MRR growing, CAC low, you hate board pace with a young family
Seed raise Clear multi-payer wedge + design partners + “compliance AI” narrative
SBIR Phase I Deep-tech ceiling + non-dilutive cash; entity + proposal ready
Hybrid SaaS cash-flow company + separate SBIR entity (common; needs clean IP/ethics structure)

Phase 4 — Separation readiness (Jul–Oct 2027)

8. Weekly system while still in the Navy

Not full-time founder yet. Act like a part-time CEO with a day-job subsidy.

Block Hours Work
Deep build 8–12 Only floor product
Customer / sales 3–5 Calls, demos, onboarding
Ceiling track 2–4 SBIR research / relationships / proposal skeleton
Admin / finance 1 Runway, entity, ethics notes
Cap ~15–20 Everything else is a distraction
Kill rule: if a side project doesn't feed Floor or Ceiling in 30 days, archive it. Quintile maintenance is fine. Another Buffett feature sprint is not.

9. Raise money and create tremendous value

Raising is not the goal. Leverage is.

Tremendous value for others, in this biography, looks like:

  1. Billers get hours back and fewer claim surprises.
  2. Operators in locked environments get automation without waiting for IT.
  3. Undersea / defense gets cheaper, quieter, more persistent sensing (if SBIR).

That's three different companies. Pick one customer face for the next year.

10. If Grok forced my hand

Given July 2026, the repos, and October 2027 separation:

  1. Floor = healthcare eligibility product for billers. Monetize with a real design partner this quarter.
  2. Ceiling prep = SBIR passive acoustics at ~20% time: entity strategy, topic search, technical advisors, Phase I outline — launch window timed to leave / ethics clearance.
  3. Mission Control / SF automation = portfolio + narrative, not a Navy product company while still in.
  4. Quintile = brand proof that polished software ships. Optional Pro tier later; not the income plan.
  5. Buffett stack = personal tool, not company.
  6. Do not take a job unless a defense company offers founder-level equity and a problem that matches the deep-tech path better than solo SBIR. Interview for signal only.
  7. Drag rental: treat as a cash-flow problem to solve in 2026, not a forever hold.

11. The uncomfortable truth

You're already capable of building more than most seed-stage teams. The risk is infinite optionality:

Venture-scale outcomes come from boring repetition on one buyer until the product is inevitable — then expanding. The next year is not “find the perfect idea.” It's make one person dependent on software you own, while the W-2 and balance sheet remove existential panic.

You already have the hard part most founders beg for: runway + skills + domain scars. The scarce resource is focus under a uniform.

12. Immediate next actions (this week)

  1. Write post-Navy monthly floor / comfort numbers with the household.
  2. Confirm one design partner for the biller tool and a pilot price.
  3. Ethics counseling note: what can be invoiced / formed / pitched while active duty.
  4. Freeze non-core repos as archive / personal.
  5. One-page: Floor product ICP + Ceiling SBIR thesis (two paragraphs each).

That's the map. Publish it so I can't pretend I didn't hear it.