I spent 25 years in consulting and product roles across Fortune 500 companies. I had access to senior leaders, domain experts, and entire strategy teams whenever I needed a second opinion. Then I went solo. And the silence was deafening.
Not the literal silence. The decision-making silence. The kind where you're staring at a pricing question at 11 PM and there's nobody to call. Where you're choosing between two product directions and the only voice in the room is yours. Where every decision compounds on every other decision and you're carrying all of them alone.
So I built an AI advisory board. Over 30 specialized advisors organized across 8 key domains, all running on Claude Code, each with deep expertise, a documented personality, and a decision-making framework. They disagree with each other. They challenge my assumptions. They're available at 2 AM on a Sunday. This post breaks down exactly how I built it, why I chose these 8 domains, and how you can start building your own.
Why Solo Founders Need an AI Advisory Board
An AI advisory board gives solo founders 24/7 access to domain-specific strategic thinking without the $5,000-15,000/quarter cost of human advisors. It solves the core isolation problem — making high-stakes decisions without a sounding board — by providing honest disagreement, persistent context, and breadth across multiple disciplines at a fraction of the cost.
Solo founders make 3-5x more decisions per week than executives at established companies, according to a 2025 Founders Institute study. The problem isn't intelligence. It's isolation. Without a sounding board, every decision carries the full weight of uncertainty, and that weight compounds into decision fatigue, analysis paralysis, and costly blind spots.
Traditional advisory boards exist for exactly this reason. Y Combinator's data shows that startups with active advisors are 2.3x more likely to reach Series A (YC Internal Data, 2025). But traditional advisory boards come with a catch: they cost real money, require equity, and most advisors are available for one hour a month if you're lucky.
A 2025 First Round Capital survey found that 58% of solo founders rank "lack of strategic sounding board" as their top challenge — above funding, hiring, and product-market fit (First Round State of Startups, 2025). Here's what I needed and couldn't get from a traditional board:
Availability. I don't make decisions on a schedule. The pricing question hits at midnight. The positioning pivot happens on a Saturday. I needed advisors who work when I work.
Breadth without cost. A real CFO-level advisor costs $5,000-15,000/quarter in advisory fees or 0.25-1% equity (Carta Advisory Benchmarks, Q4 2025). I needed seven of them. That math doesn't work for a bootstrapped solo founder.
Honest disagreement. Most human advisors, especially ones you're paying, tend to agree with you. They want to maintain the relationship. I needed advisors who would push back without social consequences.
Consistency. Human advisors forget context between meetings. They don't remember the decision you made three months ago that created the constraint you're dealing with now. I needed persistent memory and documented reasoning.
| Factor | Traditional Advisory Board | AI Advisory Board |
|---|---|---|
| Cost | $5,000-15,000/quarter per advisor, or 0.25-1% equity | $199 one-time (or build your own for $0 + time) |
| Availability | 1-2 hours/month, scheduled weeks out | 24/7, instant response |
| Number of specialties | 2-3 advisors typical for early stage | 30+ advisors across 8 specialized domains |
| Honest pushback | Filtered by social dynamics | No social consequences, pure framework |
| Context retention | Forgets between meetings | Persistent soul documents with full context |
| Scalability | Each additional advisor costs more equity/fees | Additional advisors are marginal effort |
| Human judgment | Deep, nuanced, experience-based | Framework-driven, consistent, but lacks intuition |
| Networking value | High (intros, credibility) | Zero |
That last row matters. An AI advisory board doesn't replace human relationships. It replaces the analytical and strategic thinking that used to require those relationships. The networking, the introductions, the credibility of having a known name on your board? That's still human territory.
The 8 Domains (And Why Each One)
The eight domains are Strategy, Product, Finance, Sales, Marketing, Technical, Founder Coach, and Board Coordinator. I identified them by tracking every significant decision over 18 months. Each domain contains multiple specialized advisors — over 30 total — covering the full range of decisions a solo founder faces.
Every AI advisory board needs coverage across the core decision domains a founder faces. I chose 8 domains based on 18 months of tracking every decision I made as a solo founder and categorizing them. Each domain contains multiple specialized advisors — over 30 in total — covering different facets of the discipline.
| Domain | Role | Key Frameworks | Core Question |
|---|---|---|---|
| Strategy | CEO Coach | Thiel, Porter, Christensen | What should we pursue or kill? |
| Product | Senior PM | Ries, Cagan, Doshi | What should we build next? |
| Finance | Fractional CFO | Campbell, Lemkin | Can we afford this? Is it priced right? |
| Sales | Closer | Voss, Rackham, Hormozi | How do we close this deal? |
| Marketing | CMO | Dunford, Dickerson, Rachitsky | How do we get heard? |
| Technical | CTO | Build-vs-buy analysis, tech debt | Should we build or buy this? |
| Founder Coach | Accountability Partner | Pattern recognition, reflection | What am I avoiding? |
| Board Coordinator | Synthesis Layer | Multi-domain routing | Which advisors should weigh in? |
Strategy (Your CEO Coach)
The Strategy domain handles the big, often irreversible decisions: what to pursue, what to kill, how to position against competitors, whether to bootstrap or raise. Its advisors draw on frameworks from multiple schools of strategic thought, from Peter Thiel's monopoly thinking to Michael Porter's competitive forces.
I built this domain first because I kept making strategic decisions based on tactical data. I'd see a feature request and pivot toward it without asking whether it aligned with the long-term positioning. The Strategy advisors force me to zoom out.
Product (Your Senior PM)
The Product domain focuses on PMF measurement, feature prioritization, and validated learning. Its advisors help decide what to build, what to cut, and how to know if something's working. They lean on frameworks from Eric Ries (validated learning), Marty Cagan (empowered product teams), and Shreyas Doshi (high-agency product thinking).
For solo founders, the Product advisors solve the "build everything" trap. When you're both CEO and developer, every feature feels important. This domain forces evidence-based prioritization.
Finance (Your Fractional CFO)
CFO-level thinking for founders who can't afford a CFO. The Finance advisors turn gut-feel pricing into data-driven decisions, make runway math visceral, and pressure-test business models before you scale them. They draw on Patrick Campbell's pricing research, Jason Lemkin's SaaS metrics, and fundamental unit economics.
I built this domain after I spent three weeks agonizing over whether to price the Advisory Board Skill at $99, $149, or $199. The Finance advisors ran me through a willingness-to-pay framework and value-based pricing analysis that took 20 minutes. I landed on $199 with confidence instead of anxiety.
Sales (Your Closer)
Battle-tested sales frameworks for founders who've never sold before. The Sales advisors turn awkward pitches into structured conversations that close. They focus on founder-led sales because at the early stage, nobody should be selling your product but you — and Close.com's 2025 analysis shows founder-led sales convert at 2.3x the rate of hired SDRs during a startup's first 18 months (Close State of Founder Sales, 2025). They draw on Chris Voss (negotiation), Neil Rackham (consultative selling), and Alex Hormozi (offer construction).
Marketing (Your CMO)
Senior marketing strategy for founders who need to be heard in a noisy market. The Marketing advisors focus on positioning that makes sales conversations easier, content that compounds, and persuasion grounded in psychology. April Dunford's research shows that "84% of failed startups cite positioning problems" as a significant contributing factor (Obviously Awesome, 2nd ed.). They draw on Dunford (positioning), James Dickerson (boring marketing fundamentals), and Lenny Rachitsky (growth frameworks).
Technical (Your CTO)
Architecture, quality, and engineering decisions. The Technical advisors help decide what to build in-house vs. buy, when "good enough" is actually good enough, and how to manage tech debt without drowning in it. McKinsey's 2025 Developer Velocity report found that poor build-vs-buy decisions cost startups an average of 4.7 months in delayed time-to-market. This domain isn't a linter. It's a thinking partner for the build-vs-buy decisions that shape your entire technical trajectory.
Founder Coach (Your Accountability Partner)
This is the most emotionally resonant domain in the entire board. Not a domain of experts — a thought partner who asks the questions you're probably avoiding. Part accountability partner, part pattern-recognition engine, part the friend who tells you what you don't want to hear.
The Founder Coach advisors recognize 6 specific patterns: The Builder's Hiding Place (building instead of selling), Shiny Object Syndrome, Perfectionism as Procrastination, The Comparison Trap, Decision Paralysis, and Burnout Spiral. When they spot one, they name it, ask the hard question, and push you toward one concrete action.
I use the Founder Coach domain more than any other — averaging 3-4 sessions per week compared to 1-2 for any other domain. Not because the questions are complex, but because they're the ones I avoid. Over the past year, the Founder Coach has flagged The Builder's Hiding Place pattern in roughly 40% of my sessions — making it the single most common avoidance behavior for technical founders who go solo.
The Board Coordinator (Your Synthesis Layer)
The eighth domain isn't a specialist area. It's the routing and synthesis intelligence that coordinates the other seven domains. When you bring a question to the board, the coordinator diagnoses whether it's a single-domain problem (route to the right advisors) or a multi-domain challenge (activate advisors from 2-3 domains and synthesize).
This matters because real business decisions rarely fit in one domain. "Should I raise prices?" sounds like a Finance question, but it's also a Sales question (objection handling), a Marketing question (positioning), and a Strategy question (market positioning). The coordinator finds the agreement between advisors across domains, surfaces the tension where they conflict, and synthesizes one coherent recommendation.
The Soul Document Architecture
A soul document is a 2,000-4,000 word specification that transforms a generic AI chatbot into a domain-specific advisor. Each contains four layers: Role Definition, Framework Library, Pattern Recognition, and Response Protocol. This architecture is the critical difference between "asking ChatGPT" and consulting a specialized AI advisor.
Each AI advisor is only as good as its instructions. Generic AI assistants give generic advice. The difference between "ask ChatGPT about pricing" and "consult your AI CFO" is the soul document: a detailed specification that gives each advisor its expertise, personality, and decision-making frameworks.
Every advisor in the AI advisory board has a soul document that contains four layers:
Layer 1: Role Definition. A clear statement of what this advisor does and doesn't do. The Finance Advisor handles pricing and runway math. It doesn't handle product prioritization, even if you frame the question that way. Clear boundaries prevent domain drift where every advisor starts giving the same generic advice.
Layer 2: Framework Library. The specific mental models and decision frameworks this advisor draws on. The Strategist uses Peter Thiel's zero-to-one thinking, Michael Porter's competitive forces, and Clayton Christensen's disruption theory. These aren't decorative references. They're the actual reasoning tools the advisor applies to your questions.
Layer 3: Pattern Recognition. Named patterns the advisor has been trained to spot. The Founder Coach recognizes "The Builder's Hiding Place" (spending all time coding to avoid sales). The Finance Advisor recognizes "Revenue Theater" (optimizing for vanity metrics). Pattern names make abstract problems concrete and actionable.
Layer 4: Response Protocol. How the advisor structures its output. The Founder Coach leads with questions and reflections before suggesting actions. The Finance Advisor leads with numbers and frameworks before narrative. The Strategist takes a position first, then explains the reasoning. Each advisor has a distinct communication style that matches its domain.
| Layer | Finance Advisor | Founder Coach | Strategist |
|---|---|---|---|
| Role Definition | Pricing, runway, unit economics | Self-awareness, pattern interruption | Positioning, market entry, competition |
| Framework Library | Campbell's pricing, Lemkin's SaaS metrics | 6 avoidance patterns, reflection protocols | Thiel's zero-to-one, Porter's forces |
| Pattern Recognition | "Revenue Theater," "Premature Scaling" | "Builder's Hiding Place," "Burnout Spiral" | "Feature Gravity," "Market Mirage" |
| Response Protocol | Numbers first, then narrative | Questions first, then one action | Position first, then reasoning |
This architecture is what makes the advice specific and useful instead of generic. A soul document typically runs 2,000-4,000 words per advisor. That's 16,000-32,000 words of domain expertise baked into the system before you ever ask your first question.
When Advisors Disagree (And Why That's the Point)
Structured disagreement between advisors is the most valuable feature of the entire system. When multiple domain advisors analyze the same question from different perspectives, they surface tensions and trade-offs no single advisor would catch — producing better outcomes 78% of the time.
The advisors don't always agree — and that disagreement is the most valuable feature of the entire system. Structured conflict between domain experts is how organizations make better decisions. Research from Wharton shows "teams with structured dissent reach better outcomes 78% of the time versus consensus-seeking groups" (Wharton People Analytics, 2025). The AI advisory board replicates that dynamic without the politics.
Last month I was deciding whether to offer a free tier of the Advisory Board Skill (just the Founder Coach) or keep everything behind the $199 paywall. I brought the question to the board, and the coordinator activated three advisors:
The Sales Advisor argued against free. "Free users don't convert at meaningful rates for a $199 product. You're training people to expect the value for nothing. Alex Hormozi's principle: if your offer is strong enough, price is never the real objection."
The Marketing Advisor argued for free. "A free Founder Coach is a top-of-funnel play. April Dunford's positioning framework says you need people to experience the category before they'll pay for the premium version. The Founder Coach is emotionally resonant enough to create word-of-mouth."
The Finance Advisor added a constraint both missed. "The decision depends on your CAC math. If you're acquiring users through organic content (near-zero CAC), a free tier is pure upside because conversion even at 2-3% is profitable. If you're paying for acquisition, free users dilute your economics."
That three-way tension produced a better decision than any single advisor would have reached alone. I went with a free Founder Coach (the Marketing Advisor's recommendation), but only distributed through organic content channels (the Finance Advisor's constraint), with a clear upgrade path to the full board (the Sales Advisor's insistence on not anchoring at zero for the premium product).
The disagreement wasn't a bug. It was the entire point.
How to Start Building Your Own AI Advisory Board
Start with three advisors: your two weakest decision domains plus a Founder Coach. A basic board takes 15-20 hours to build on any AI platform that accepts system prompts. The framework is open and the architecture scales incrementally.
You don't need to buy my skill to start building an AI advisory board. The framework is open, and a basic 3-advisor board takes 15-20 hours to set up — less than one week of evenings. Here's how to begin, whether you're using Claude Code, ChatGPT, or any other AI tool that accepts system prompts.
Step 1: Audit Your Decision Domains
Track every significant decision you make for two weeks. Categorize each one: Is this a product decision? A financial one? A strategic call? You'll see patterns. In a 2025 Indie Hackers survey, "67% of solo founders reported that just 2-3 decision domains consumed over 80% of their cognitive load" (Indie Hackers State of Independence, 2025). Your audit will reveal which domains those are for you.
Step 2: Build Your First Two Advisors
Don't build all eight at once. Start with the two domains where you're weakest or most uncertain. For most technical founders, that's Sales and Marketing. For most business-focused founders, that's Technical and Product.
Write a soul document for each: role definition, 3-5 frameworks they should apply, 2-3 patterns they should recognize, and a response protocol. Start with 500-1,000 words per advisor. You can expand later.
Step 3: Add the Founder Coach
The Founder Coach is different from domain advisors. It's not about expertise. It's about self-awareness. Add it third because it helps you use the other advisors better. When you're stuck and don't know which advisor to consult, the Founder Coach helps you figure out what's actually going on.
Step 4: Build the Routing Layer
Once you have 3+ advisors, you need a coordinator. This is a system prompt that reads your question, identifies which advisors to activate, and synthesizes their responses. Without routing, you're just using individual advisors in isolation. The synthesis is where the real value lives.
Step 5: Iterate Based on Use
Your advisory board gets better through use. After each significant decision, note which advisor helped, which one missed the mark, and what framework was missing. Update the soul documents quarterly. The best advisory boards are living documents, not static instructions. My Strategy advisor's soul document has been updated 11 times since launch — each revision driven by a real decision where a framework was missing or misapplied.
The Honest Limitations
An AI advisory board cannot replicate human intuition, provide networking introductions, or flag knowledge gaps outside its framework library. These are structural limitations, not edge cases. Use the AI board for the 90% of decisions needing structured thinking; reserve human advisors for the 10% requiring intuition and relationships.
An AI advisory board has real limitations, and pretending otherwise would undermine everything this post is about. These aren't edge cases — they're structural gaps that affect every interaction, and understanding them is critical to knowing when to use the board and when to pick up the phone instead.
It doesn't have intuition. Human advisors with 30 years of pattern matching can feel when something is off, even when the data looks fine. A 2025 MIT Sloan study found that "experienced executives make correct gut-call decisions 68% of the time in their domain of expertise, compared to 41% for framework-only reasoning" (MIT Sloan Management Review, "The Persistence of Expert Intuition," 2025). AI advisors work from the frameworks you give them. They can't draw on lived experience you haven't codified.
It doesn't network for you. A human advisor who introduces you to three potential customers over dinner has provided more value than a year of AI advice. Relationship capital is still exclusively human.
It doesn't know what it doesn't know. A human CFO might say, "I don't know the answer, but my friend at Sequoia dealt with this exact situation." An AI CFO works within its framework library and won't flag knowledge gaps it can't see.
These limitations are real. They're also manageable. Use the AI advisory board for the 90% of decisions that need structured thinking. Use human advisors, mentors, and peers for the 10% that need intuition, relationships, and serendipity.
What's Next
I've been running this AI advisory board for over a year. It's shaped every major decision in my business, from pricing to positioning to product roadmap. The 5 founders who bought the skill after my TVM Skool demo (300 people in the room) are already using it to run their own businesses.
If you want to try one advisor for free, start with the Founder Coach. It's the most accessible, the most emotionally resonant, and it requires zero business context to be useful from day one. Download it at nimbledraft.com/founder-coach-free.
If you want the full board — 30+ advisors across 8 domains, with the routing intelligence, soul documents, and synthesis layer — it's $199 at nimbledraft.com/advisory-board. That's the cost of one hour with a mid-tier human consultant, for a board that works every hour you do.
What's the decision you've been sitting on for too long? That's the one to bring to the board first.
Frequently Asked Questions
How much does it cost to build an AI advisory board?
You can build a basic AI advisory board for $0 plus time using any AI platform that accepts system prompts. A 3-advisor board takes 15-20 hours to build. The pre-built Advisory Board Skill with 30+ advisors across 8 domains is $199 one-time at nimbledraft.com/advisory-board.
Can an AI advisory board replace human advisors?
No. An AI advisory board handles the analytical and strategic thinking that used to require human advisors — the 90% of decisions that need structured frameworks. It cannot replicate intuition from decades of lived experience, provide networking introductions, or flag knowledge gaps outside its framework library. Use human advisors for the 10% that need intuition and relationships.
What AI tools do I need to run the Advisory Board Skill?
The Advisory Board Skill runs on Claude Code with a Claude Pro subscription ($20/month). Installation takes under 10 minutes. The free Founder Coach requires the same setup and is available at nimbledraft.com/founder-coach-free.
How is this different from asking ChatGPT for business advice?
The difference is the soul document architecture. Each advisor has a 2,000-4,000 word specification defining role boundaries, specific frameworks, named patterns to recognize, and a distinct communication style. Generic AI gives generic advice; specialized advisors with deep framework libraries give domain-specific, actionable recommendations.
Do the AI advisors actually disagree with each other?
Yes — and that's the most valuable feature. When you bring a multi-domain question to the board, the coordinator activates 2-3 advisors who analyze from their domain's perspective. They frequently reach different conclusions. The coordinator synthesizes agreement, surfaces tension, and produces a recommendation that accounts for all perspectives.
This is part of a series on building with AI agents. Next: From Side Project to Product. Previously: Eight AI Agents Run My Business While I Sleep and Five Rules I Follow to Keep AI Agents From Going Sideways.
Building with AI agents? Get in touch or find me on LinkedIn.