This document was generated with the help of Claude Opus 4.6 reading from aswritten.ai's collective memory. Every claim traces back to a specific person, context, and decision. The citations are not decoration — they are the product working.

The Opportunity

Every organization deploying AI is building their own approach to organizational context — knowledge bases, project instructions, documentation pipelines, system integrations. The tools for getting context into AI have never been better. But they can only access knowledge that's been written down. And most of what an organization knows has never been written down.[1]

As senior staff work with AI, their tools accumulate months of context about how the organization operates — but they're missing the opportunity to extract the decisions behind their work. The reasoning, the methodology, the "why we did it this way" stays locked in their heads. When they're busy, everyone waits. When they leave, everything they knew leaves with them.[2]

OpenAI validated the category in early 2026 when they invested in Frontier to build a shared context layer for enterprise AI agents. But Frontier assumes knowledge already exists in enterprise systems. It doesn't. Someone has to extract it first.[3]

Extracting undocumented knowledge is the opportunity. Every company deploying AI agents needs organizational memory. The market is every organization with expertise worth preserving, and every task done by someone who would otherwise have to ask for help.[4]

What I Built

aswritten.ai solves the extraction problem. Through structured conversational interviews, we draw undocumented knowledge out of the people who hold it — capturing not just decisions but who made them, why, and what alternatives were considered. That knowledge becomes versioned organizational memory that any AI tool across the organization can query.[5]

Every piece of knowledge carries conviction and provenance. A passing idea is marked differently from a settled architectural decision. Every claim traces back to the person who made it, the context they made it in, and what it replaced. This isn't a wiki where someone typed something once and it slowly goes stale — it's a living worldview where knowledge has weight, attribution, and history.[6]

Other approaches to organizational context give AI more information. We give it direction — the ontology ensures AI doesn't just know more facts about the organization, it reasons with the organization's worldview.[7]

A new organization signs up and connects to their existing codebase. A consultant-led onboarding session interviews senior staff, tracing decisions to the people who made them — but the system also captures knowledge as decisions happen in calls, working sessions, and day-to-day AI interactions. The organizational worldview grows with every conversation.[8]

That worldview is versioned and branchable. Different teams or projects can branch their perspective the same way developers branch code — proposing a new direction, reviewing how it shifts the worldview, and building consensus before merging it into shared reality. When the worldview updates, downstream pipelines respond automatically: regenerating onboarding docs, sales decks, and board updates, notifying stakeholders about changes filtered by and personalized to their role.[9]

The system also captures organizational style and voice — so AI outputs don't just know what the company decided, they sound like the company. Individual team members build personal context that composes with the shared organizational worldview, so every AI interaction reflects both the person and the organization they belong to.[10]

The result is that anyone — including interns, new hires, and staff with no product training — can work with AI that already knows how the company operates, how it communicates, and why it made the choices it did.[11]

The product is live and in use across multiple AI platforms — not locked to any single vendor. This document was generated from aswritten's own collective memory.

Why Me

I started programming the flocking of birds in 2011[12] and saw the simple rules followed by each bird — to not collide with their fellow, to not stray too far, and to align with others — produced incredibly complex and beautiful emergent behavior, impossible to predict from the top down.[13]

I spent years working with language as code, and with LLMs in 2019 (then GPT-2) I saw that similar to the steering vectors that described the flight of those birds, that these language machines designed to predict the next word also had a kind of velocity, one steered by narrative — the directional vector of meaning beneath statements that say something far beyond their single sentence.[14]

Aswritten is my 5th company after a digital agency, a venture funded startup that was later acquihired, an arts collective and events company, and a training program in narrative strategy for entrepreneurs.[15] I taught myself to program in the 3rd grade and spent the first decade of my professional years writing software in data visualization, natural language processing, and mapping novel user experience across pre-existing technology.[16] Since 2015 I have mixed software consulting with entrepreneurship, and in 2022 formalized a narrative strategy consultancy.[17]

As Chief Strategist at Vouch.io, I developed a manual process for steering LLMs with narrative.[18] Instead of training new models, I treated the model like a computer and a narrative architecture like a program — an interconnected web of narratives that define a way of thinking.[19] I used this to bootstrap any AI to instantly think like the company: to create content, test ideas, even write code. Instead of defaulting to the average of the Internet, the underlying model was now flying in the direction of the company.[20]

I left Vouch in September 2025 to automate the process and built Aswritten.ai — collective memory for AI.[21] Every conversation with a coworker, every piece of client feedback, every strategic argument, every engineering decision — these are all course corrections that currently evaporate. Aswritten catches the heading change before it's lost and breaks it down into a narrative architecture of steering statements that point toward where we're going, not just where we are today.[22]

I know this product works because I use it myself everyday as I've built this platform solo in 6 months, AI gave me the speed and Aswritten gave AI the controllable direction.[23] It's what I needed for my team at Vouch, and it's what I hear time and time again is what's needed by my advisors, beta testers, and sales conversations as I move closer to product market fit.[24]

LLMs do feel like flying. But they have turned our starting line from the blank page to the complete and misguided draft.[25] I built Aswritten to feather our arrows, that our first attempt may fly true.[26]

Our direction, as written.

Traction & Proof

The product is in production and under load.[27] In the last 30 days: 328 knowledge extractions run, 1,673 automations triggered from worldview changes, and 250M tokens passed through the system.

Since launching in September 2025,[28] I've run 3-5+ discovery calls, advisor sessions, and product demos every week[29] — shaping the product through continuous feedback across enterprise software, AI-native dev shops, education, healthcare, design, enterprise consulting, and solo founders.[30] Seven organizations are now actively building collective memory:

  • Martin Kess — Former CTO of Fractal Labs (AI-focused incubator), founder of PurpleFish.ai (~$400K revenue voice agent company, 50-100 PRs/day).[31] Independently validated that planning — not coding — is the bottleneck for AI-native teams.[32] Integrating the Aswritten MCP server into his new open-source project, factoryfactory.ai.[33]
  • Mike Sackton — Chief Architect, Escher Group (enterprise postal software, 25 years).[34] Validated the core senior-expert extraction pattern.[35] Identified QA onboarding as a "killer sales demo."[36] His team confirmed Aswritten works across Claude, GitHub Copilot, and Codex — not locked to any single AI vendor.[37]
  • Roger Beaman — Recently exited founder (acqui-hired), building SortaRich.com entirely via AI-driven development.[38] His onboarding validated the sales-as-delivery motion live — the call itself produced his collective memory.[39] Introduced Martin Kess to Aswritten.[40]
  • Daniel Donoghue — Designer and creative director, former Vouch.[41] Shifted from ChatGPT to Claude and uses Claude CLI connected to Figma via MCP for design-to-code workflows.[42] Validated that branching worldviews resonate with real user pain.[43]
  • Clay Harper — Enterprise consultant and designer specializing in systems integration and data-driven tools for Fortune 1000 companies, major defense contractors, and sovereign governments.
  • Peter Caron — Small business owner pioneering AI automations in healthcare and patient outreach.[44]
  • Taylor Passmore — Founder of Bluebird Learning Partners, education consultant serving multiple school districts and organizations.[45] Validates the multi-client use case — one shared methodology with separate collective memory for each client, so every AI interaction reflects both Bluebird's approach and the specific school's context.[46]

Advisors: Lucinda Treat (CEO, Crooked Media; formerly chief counsel, Vice Media), Anthony Maley (former CEO, Vouch — bootstrapped to $5M ARR shipping decentralized identity to Toyota and Lexus), Bill Cromie (formerly Managing Director, Robin Hood Labs; CTO, Vouch).[47]

OpenAI validated the category when they invested in Frontier in early 2026 to build a shared context layer for enterprise AI agents.[48] Frontier assumes knowledge already exists in enterprise systems.[49] Aswritten solves the upstream problem — extracting the knowledge that was never written down.[50]

Every sales call builds the prospect's collective memory before they decide to buy.[51] The sales motion and the product delivery are the same activity[52] — which means the pipeline is also product validation.[53]

Business Model

Aswritten's revenue model is built on a structural insight: the work required to onboard a customer is the same work that delivers the product's core value.[54] A consultant-led session gathers the raw material — interviews with senior staff, existing documents, decision history.[55] The product already handles what happens next: extracting that material into a narrative architecture of conviction-weighted, provenanced knowledge.[56] That engine is built and running.[57] The onboarding work is filling the tank, not building the engine.[58]

By the end of an engagement, the customer has a working organizational worldview — and every AI tool in their organization can query it.[59] There is no gap between sales and delivery.[60] The setup work is the value.[61]

The product is priced across four tiers that map to a natural expansion path:

  • Free — Open source repos. Collective memory is public.
  • Individual ($80/month) — Private repo, single user. No collaborators.
  • Organization ($400/month) — Unlimited repos, up to 100 users. Includes knowledge extraction package.
  • Enterprise ($5,000/month + $10K setup) — No repo or user limits. Setup fee covers knowledge extraction engagement with multiple expansion phases for automations and cross-team integration.[62]

All tiers control token spend by limiting usage per organization and allowing customers to bring their own AI vendor API key through OpenRouter.[63] This keeps margins predictable and gives enterprise customers control over their own spend.[64]

The expansion motion is organic. A developer discovers Aswritten through an open source project and connects their own API key. They start a private project and move to Individual. They invite a teammate and upgrade to Organization. As collective memory compounds across their team, they want bespoke automations and cross-team integration — Enterprise.[65] Each step up is driven by the data moat the customer has already built.[66]

Unit economics favor the model at scale. Token cost per customer runs $20-60/month, yielding gross margins of 85-95% on the product side. Consulting engagements are margin-positive from the first call.

The central question for any investor is how this scales beyond founder-led intake. The extraction engine — the part that turns raw material into queryable narrative architecture — already works.[67] What's manual today is the gathering: conducting interviews, identifying the right people, pulling the right documents.[68] The AI-led version of that intake already exists — you can connect to Aswritten today and an AI will walk you through extracting your first memories.[69] The flow and prompts need iteration to match the quality and depth of a consultant-led session, and that iteration comes from real usage and studying the transcripts of human-led sessions to extract the conversational patterns that make intake work well.[70] Every consultant engagement produces the material to make the automated intake better.[71] The company is progressively automating the gathering, not the extraction — the extraction is already the product.[72]

Go-To-Market

The sales call is the product demo is the delivery.[73] A prospect gets on a call, and within an hour their senior knowledge has been gathered and extracted into a working collective memory.[74] Their AI tools produce output that is noticeably more aligned to organizational direction than anything possible before the call.[75] They leave with access to that repo on a limited token budget — enough to experience the difference, not enough to run their organization on it.

From there, the data moat does the selling.[76] Every memory the prospect saves, every decision they capture, every team member who queries the worldview — all of it builds organizational context that only works through Aswritten's tools.[77] Introspect, annotate, compile, story generation — these are the capabilities that make the accumulated knowledge actionable.[78] The prospect isn't evaluating a demo anymore.[79] They're using a product that already knows their business, and they need to pay to unlock its full power.[80]

Two channels feed the pipeline:[81]

Direct network — Scarlet's 3-5 weekly calls convert into team and organization subscriptions.[82] The sales cycle is short because the call itself is the proof.[83] The beta cohort came through this channel — advisor introductions, founder networks, existing relationships.[84] This is the primary channel through the F&F raise period.[85]

Enterprise pipeline — Tony Maley's enterprise relationships and VC network open doors to larger organizations where the engagement starts with a $10K knowledge extraction setup and expands into cross-team integration and bespoke automations.[86] Longer cycle, higher annual contract value.[87] This channel activates as Tony's involvement formalizes.[88]

Early customers come through high-touch consultant-led gathering.[89] As the automated intake improves — refined through real usage and the patterns extracted from human-led session transcripts — the sales motion shifts from "get on a call with Scarlet" to "connect your repo and start talking to the AI."[90] The same expansion path that grows individual accounts from free to enterprise also grows the company from founder-led sales to product-led growth.[91]

The Ask

Raising $275K from friends and family on a post-money SAFE using standard YC terms via Clerky.[92] This funds the company through first enterprise pilots and initial revenue — the proof points needed to raise venture capital at the end of fall 2026.[93]

This is not a gift.[94] It's an investment in a working product with traction, offered on standard terms with standard protections.[95] The thesis: F&F money turns a solo-built product with a seven-org beta into a company with paying enterprise customers, case studies, and unit economics that make a venture pitch.[96]

Use of Funds & Phases

Monthly Duration Total
Founder salary (net $9K, 1.6x gross-up) $14,400 10 months $144,000
Full-time engineer (1099) $15,000 7 months $105,000
Production (AI API, hosting) $2,000 10 months $20,000

Spring (March–May): Betas and initial pilots.[97] The seven-org beta converts into paying customers on organization subscriptions. The automated onboarding flow continues to improve through real usage.[98] Initial pilot conversations with enterprise prospects begin.[99] Revenue starts.[100]

Summer (June–August): First enterprise pilots.[101] A full-time engineer joins to harden the product and build infrastructure for paying enterprise customers.[102] First enterprise pilots close with $10K setup fees.[103] The sales-as-delivery motion produces case studies with every engagement.[104]

Fall (September–November): Case studies and seed prep.[105] Enterprise pilots produce the evidence a venture pitch requires — named customers, retention data, unit economics, expansion revenue.[106] The automated intake continues to improve, reducing founder dependency on every onboarding.[107] Product-led signups through free and individual tiers establish a bottom-up growth channel alongside the consultant-led enterprise motion.

Winter (December onward): Raise venture or sustain on revenue.[108] If pilots convert and expand, the company raises a seed round to scale — more engineers, more automation, broader market.[109] If the venture path is not ready, consulting revenue through PnD sustains the company while the product continues to compound.[110] Either path is forward.

Risk & Downside

The honest risks:

Solo founder.[111] Capacity maxes at roughly 1.5 concurrent engagements today.[112] The engineer hire in summer extends this, and the progressive automation of intake extends it further, but until then the company scales at the pace of one person's calendar.[113]

Market timing.[114] Tony's read is a two-year window before legal and corporate pressure reshapes the AI landscape — pattern-matched to the early internet boom.[115] Moving fast matters more than moving perfectly.

Pre-revenue.[116] The product is in production and beta users are active, but no one has paid yet.[117] Spring is when that changes. If it doesn't, the thesis needs revisiting.

Technical debt.[118] The product was built fast by one person.[119] An engineer joining in summer inherits a codebase that prioritized shipping over architecture.[120] This is normal for the stage but it's real work.

The downside is not zero — but it's bounded.[121] In the worst case, the product doesn't find market fit and SIC winds down.[122] SAFE investors get 1x back on dissolution per standard terms. What remains is a senior AI consultant with a proprietary system, seven organizational case studies, and deep deployment knowledge.[123] The PnD consulting practice is stronger than it was before SIC existed.[124] The floor is a better consulting business, not starting over.[125]

Open Questions

These are the things I'm still figuring out. Including them here signals maturity, not weakness — they're the right questions for this stage.

  • What's the right individual tier price point? $80/month is the current plan but hasn't been tested with real buyers.[126]
  • At what point does enterprise compliance (SOC 2, GDPR, data sovereignty) become a gate rather than a nice-to-have?[127] Mike Sackton confirmed it's a real adoption gate for enterprise, but the timing is an open question.[128]
  • When does the second extractor hire become necessary? The progressive automation buys time, but at some point the human-led intake needs to be a team, not a person.[129]
  • What's the right balance between placing the product with high-visibility organizations for free to build leverage, versus charging early to establish pricing?[130] Both have strategic value.[131]
  • Is the operations persona — Series A ops people managing fragmented context across Notion, Slack, and a dozen AI tools — a stronger entry point than the developer persona?[132] Martin Kess and Billy Sylvester both pointed here.[133]
  • Can we delay part of the raise until we have 10 Team subscriptions?
  • How many concurrent organizations can we onboard and do knowledge extraction for at once?

How to Read the Citations in This Document

This document is a product demo.

Every footnote traces back to a specific "Memory" stored in our git repository. These memories are compiled into a "Worldview" — a structured graph of everything we know, believe, and have decided.

  • Source: Who said it (e.g., Founder, Advisor, Customer).
  • Conviction: How "settled" the idea is (Foundation → Boulder → Stake → Notion).
  • Provenance: The specific transaction in our knowledge graph.

When the business changes — a new customer, a pricing shift, a strategic pivot — we commit a new memory, and this brief regenerates. The diff between this version and the next is the literal record of our company's evolution.

  1. "The faster AI development moves, the more critical it is to maintain consensus on direction. Getting Claude to write a feature is easy; getting it to interact correctly with other features requires shared context." Foundation conviction. (Scarlet Dame, architecture sessions) ↩︎
  2. Mike Sackton at Escher Group (25 years as Chief Architect) is one of a small group who "know everything" about the product. The constant joke: "if only we could get AI to duplicate Mike's brain." This pattern repeats in every organization with senior expertise. Boulder conviction. (Mike Sackton, beta onboarding session) ↩︎
  3. OpenAI's Frontier investment validated shared organizational context for AI agents as an enterprise requirement. aswritten's differentiation: upstream of Frontier, solving the extraction problem that agent platforms do not address. Boulder conviction. (Scarlet Dame, Frontier research; competitive analysis) ↩︎
  4. Tony's prediction: "You've got a two year window" before legal and corporate pressure reshapes the AI landscape — pattern-matched to the early internet boom. Billy Sylvester validated the ICP at 5–150 person companies, $10–50M revenue. Boulder conviction. (Tony Maley, co-founder sessions; Billy Sylvester, advisor call) ↩︎
  5. The distinction between documentation and collective memory is foundational: documentation captures static artifacts; collective memory captures perspectives, decisions, and their underlying rationale as a living worldview. Boulder conviction. (Scarlet Dame, positioning sessions) ↩︎
  6. Conviction levels track how settled knowledge is: from Notion (emerging idea, easily moved) through Stake (planted position, needs validation) and Boulder (settled, hard to move) to Foundation (bedrock, practically immovable). Every shift in conviction traces to a specific person and context. This is orthogonal to whether the extraction has been reviewed — conviction tracks the knowledge itself. Boulder conviction. (Scarlet Dame, architecture sessions) ↩︎
  7. Declarative positioning statement: collective memory builds narrative architecture that determines how AI reasons about an organization. The steering is the product, not a side effect. Foundation conviction. (Scarlet Dame, positioning sessions; validated across four advisor/beta calls, Jan–Mar 2026) ↩︎
  8. Mike Sackton validated the core pattern: extract knowledge from Mike, ingest it, let a junior developer or CTO query it without Mike present. The first session is consultant-led; subsequent sessions the user can drive. Enterprise engagement traces decisions to named individuals with full provenance. Boulder conviction. (Mike Sackton, beta onboarding; Scarlet Dame, enterprise positioning) ↩︎
  9. Organizations shift from producing artifacts to producing worldview; artifacts become compilation targets that regenerate when the worldview changes. The branching model maps to organizational decision-making: propose a perspective shift, review how it changes downstream outputs, build consensus, merge. Boulder conviction. (Scarlet Dame, positioning sessions) ↩︎
  10. No competitor has conceptualized the personal-to-organizational compile layer. Individual context composes with shared organizational memory — the AI reflects both. Style extraction ensures outputs match organizational voice, not generic AI tone. Boulder conviction on personal graph novelty. (Scarlet Dame, architecture sessions; competitive analysis) ↩︎
  11. The creation gap means every AI tool is exactly as useful as the knowledge that's been documented — and the highest-value knowledge never has been. Collective memory eliminates this by extracting undocumented expertise and making it available to every AI tool regardless of who's using it or how long they've been at the company. Boulder conviction. (Scarlet Dame, competitive positioning reframe, Mar 2026; Martin Kess, dev-native validation, Feb 2026) ↩︎
  12. Flocking origin (2011). Foundation conviction. (Scarlet Dame, founding story through-line, Feb 2026) ↩︎
  13. Simulated flocking behavior with simple local rules producing emergent complexity; primed understanding of physics, movement of companies in ecosystems, and movement of concepts within frameworks. Foundation conviction. (Scarlet Dame, founding story through-line, Feb 2026) ↩︎
  14. Core technology insight: language models have velocity; narrative is the directional vector of meaning beneath statements. Foundation conviction. (Scarlet Dame, org architecture decisions, Feb 2026) ↩︎
  15. Each phase follows the same pattern — the physics never changed, the substrate did: generative systems → IB5K → Populous → Elephant Collective → Capitol → Vouch → aswritten. Foundation conviction. (Scarlet Dame, founding story through-line, Feb 2026) ↩︎
  16. Career arc traced from particle physics (NYU ITP) through NLP (IB5K), mass coordination (Elephant Collective), LLM theory (Clojure talk), to organizational memory (Vouch, aswritten.ai). Foundation conviction. (Scarlet Dame, founding story through-line, Feb 2026) ↩︎
  17. Matches founder's career arc and PnD consulting history. Foundation conviction. (Scarlet Dame, fundraising plan, Mar 2026) ↩︎
  18. Scarlet joined Vouch as strategist to reposition company from OEM automotive supplier into larger enterprise-facing sales organization; created what was called a narrative source of truth, done manually at the time. Foundation conviction. (Scarlet Dame, founding story through-line, Feb 2026) ↩︎
  19. Narrative architecture as interconnected web of narratives defining a way of thinking — a program installed onto model hardware. Foundation conviction. (Scarlet Dame, org architecture decisions, Feb 2026) ↩︎
  20. Created a single source of truth that allowed others to replicate AI context for content generation across the organization; instead of AI defaulting to training-set narratives, it defaulted to organizational narratives. Foundation conviction. (Scarlet Dame, founding story through-line, Feb 2026) ↩︎
  21. Created September 2025 as the automation of the Vouch process — an ingestion pipeline for arbitrary content that extracts narrative information and creates a narrative architecture. Foundation conviction. (Scarlet Dame, founding story through-line, Feb 2026) ↩︎
  22. Matches the Reviewable Worldview and Steering Vector narratives. Boulder conviction. (Scarlet Dame, sales call flow sessions, Mar 2026) ↩︎
  23. Founder built platform solo in 6 months. Foundation conviction. (Scarlet Dame, pilot plan draft, Feb 2026) ↩︎
  24. Validated by feedback from Mike Sackton, Tony Maley, and Martin Kess. Boulder conviction. (Mike Sackton, beta onboarding; Frontier competitive analysis, Feb 2026) ↩︎
  25. Contrasts LLMs moving starting line from blank page to complete-and-misguided draft. Foundation conviction. (Scarlet Dame, honest pitch manifesto, Feb 2026) ↩︎
  26. Feather arrows so first attempt goes further in the right direction; adjust heading, don't chase missed shots. Foundation conviction. (Scarlet Dame, honest pitch manifesto, Feb 2026) ↩︎
  27. OAuth discovery, PKCE flow, JWT issuance/validation, and dual auth all verified in production. Foundation conviction. (Scarlet Dame, V1 scope crystallization, Feb 2026) ↩︎
  28. Luke Vanderhart contracted for ideation and architecture on Aswritten, September 2025. Foundation conviction. (Scarlet Dame, founding story through-line, Feb 2026) ↩︎
  29. GTM process: schedule 3-5 calls with experts per week. Foundation conviction. (Scarlet Dame, consultant-led sales and pricing, Feb 2026) ↩︎
  30. Feedback loop across market segments validated through beta onboarding sessions. Foundation conviction. (Scarlet Dame, Daniel Donoghue beta onboarding, Feb 2026) ↩︎
  31. Former Google engineer, founder of voice agent company; ~$400K revenue in staffing vertical; 50-100 PRs/day AI-native workflow. Foundation conviction. (Scarlet Dame, Martin Kess discovery call, Feb 2026) ↩︎
  32. Martin validates the Feb 5 thesis: planning is the obvious next thing. Understanding the codebase is the key use case for dev-focused users. Boulder conviction. (Scarlet Dame, Martin Kess onboarding call, Feb 2026) ↩︎
  33. factoryfactory.ai plans to integrate the Aswritten MCP server. Foundation conviction. (Scarlet Dame, Martin Kess onboarding call, Feb 2026) ↩︎
  34. One of a small group who "know everything" about the product. 25 years at Escher. Foundation conviction. (Scarlet Dame, context architecture sessions, Feb 2026) ↩︎
  35. Mike Sackton validated the extract-from-senior-person pattern: extract knowledge from Mike, ingest it, give MCP server to junior developer or CTO who can query without Mike present. Boulder conviction. (Scarlet Dame, sales call flow sessions, Mar 2026) ↩︎
  36. QA onboarding scenario identified as "killer sales demo" by customer — concrete timeline, measurable outcome. Stake conviction. (Mike Sackton, advisor feedback call, Feb 2026) ↩︎
  37. MCP-agnostic architecture confirmed with production evidence: GitHub Copilot + GPT-5.2 Codex integration working. Boulder conviction. (Mike Sackton, advisor feedback call, Feb 2026) ↩︎
  38. Recently exited founder, building SortaRich.com entirely via AI-driven development. Foundation conviction. (Scarlet Dame, Roger Beaman onboarding, Feb 2026) ↩︎
  39. The call IS the first memory — stated as a foundational design principle. Foundation conviction. (Scarlet Dame, beta onboarding call process, Feb 2026) ↩︎
  40. Roger introduced Martin Kess to Scarlet Dame. Foundation conviction. (Scarlet Dame, Martin Kess onboarding call, Feb 2026) ↩︎
  41. Former designer and creative director at Vouch. Foundation conviction. (Scarlet Dame, Daniel Donoghue beta onboarding, Feb 2026) ↩︎
  42. Both Scarlet and Daniel shifted from ChatGPT to Claude; Daniel uses Claude CLI connected to Figma via MCP. Foundation conviction. (Scarlet Dame, Daniel Donoghue beta onboarding, Feb 2026) ↩︎
  43. Daniel immediately recognized branching as solving a current pain point. Foundation conviction. (Scarlet Dame, Daniel Donoghue beta onboarding, Feb 2026) ↩︎
  44. Gotham Acupuncture; small business owner + assistant. Foundation conviction. (Scarlet Dame, beta strategy reframe, Feb 2026) ↩︎
  45. Founder of Bluebird Learning Partners, education consultant. Foundation conviction. (Scarlet Dame, Taylor Passmore onboarding call, Mar 2026) ↩︎
  46. Directory inheritance enables multi-tenant hub deployment for consultants working across multiple clients. Foundation conviction. (Scarlet Dame, beta strategy reframe, Feb 2026) ↩︎
  47. Vouch alumni network and advisor relationships with deep domain context. Foundation conviction. (Scarlet Dame, Daniel Donoghue beta onboarding, Feb 2026) ↩︎
  48. Shared organizational context for AI agents validated as enterprise requirement by OpenAI's Frontier investment. Boulder conviction. (Scarlet Dame, Frontier competitive analysis, Feb 2026) ↩︎
  49. Frontier assumes knowledge already exists in enterprise systems. Foundation conviction. (Scarlet Dame, Frontier competitive analysis, Feb 2026) ↩︎
  50. Aswritten solves the upstream knowledge extraction problem that agent platforms do not address. Boulder conviction. (Scarlet Dame, Frontier competitive analysis, Feb 2026) ↩︎
  51. Every sales call builds collective memory for the prospect's repo. Foundation conviction. (Scarlet Dame, fundraising plan and sales delivery insight, Mar 2026) ↩︎
  52. The sales motion and product delivery are the same thing. Foundation conviction. (Scarlet Dame, fundraising plan and sales delivery insight, Mar 2026) ↩︎
  53. The quality of the engagement IS the product validation. Foundation conviction. (Scarlet Dame, beta strategy reframe, Feb 2026) ↩︎
  54. The sales motion and product delivery are the same thing. Foundation conviction. (Scarlet Dame, fundraising plan and sales delivery insight, Mar 2026) ↩︎
  55. Five-step beta call pattern: interview them (call IS the first memory), walk through onboarding together, upload documents, do ingestion and narrative strategy, solicit feedback in use. Foundation conviction. (Scarlet Dame, beta onboarding call process, Feb 2026) ↩︎
  56. Aswritten extracts undocumented knowledge locked in senior experts' heads through conversational AI interviews, creating version-controlled organizational worldview. Foundation conviction. (Scarlet Dame, Frontier competitive analysis, Feb 2026) ↩︎
  57. Technical infrastructure works; cognitive bootstrapping is the highest-leverage product gap. Foundation conviction. (Scarlet Dame, Martin Kess onboarding call, Feb 2026) ↩︎
  58. Onboarding mode is the highest-leverage product gap for beta — filling the tank, not building the engine. Foundation conviction. (Scarlet Dame, Martin Kess onboarding call, Feb 2026) ↩︎
  59. Mission: create a living organizational worldview that evolves through intentional memory-saving and branches like code. Foundation conviction. (Scarlet Dame, living organizational worldview, Dec 2025) ↩︎
  60. Foundation conviction. (Scarlet Dame, fundraising plan and sales delivery insight, Mar 2026) ↩︎
  61. Foundation conviction. (Scarlet Dame, fundraising plan and sales delivery insight, Mar 2026) ↩︎
  62. Enterprise pricing: ~$5K/month license plus ~$10K setup fee for upfront knowledge extraction. Stake conviction. (Scarlet Dame, weekly session and Verity demo, Mar 2026) ↩︎
  63. Enterprise users control spend via their own OpenRouter dashboard; shared key serves beta users. Stake conviction. (Scarlet Dame, V1 scope crystallization, Feb 2026) ↩︎
  64. BYOK reframed from limitation to enterprise selling point. Stake conviction. (Scarlet Dame, V1 scope crystallization, Feb 2026) ↩︎
  65. Expansion track: pilot → team license + retainer → multi-team rollout → enterprise-wide adoption. Foundation conviction. (Scarlet Dame, pilot plan draft, Feb 2026) ↩︎
  66. Data moat drives each upgrade step. Boulder conviction. (Scarlet Dame, Martin Kess discovery call, Feb 2026) ↩︎
  67. The extraction engine already works. Foundation conviction. (Scarlet Dame, Martin Kess onboarding call, Feb 2026) ↩︎
  68. What's manual is the gathering — conducting interviews, identifying the right people. Foundation conviction. (Scarlet Dame, Martin Kess onboarding call, Feb 2026) ↩︎
  69. AI-led onboarding already exists. Foundation conviction. (Scarlet Dame, Martin Kess onboarding call, Feb 2026) ↩︎
  70. Flow and prompts need iteration through real usage and studying human-led session transcripts. Foundation conviction. (Scarlet Dame, Martin Kess onboarding call, Feb 2026) ↩︎
  71. Every consultant engagement produces material to improve automated intake. Foundation conviction. (Scarlet Dame, Martin Kess onboarding call, Feb 2026) ↩︎
  72. Progressively automating the gathering, not the extraction. Foundation conviction. (Scarlet Dame, Martin Kess onboarding call, Feb 2026) ↩︎
  73. Foundation conviction. (Scarlet Dame, fundraising plan and sales delivery insight, Mar 2026) ↩︎
  74. Foundation conviction. (Scarlet Dame, fundraising plan and sales delivery insight, Mar 2026) ↩︎
  75. Foundation conviction. (Scarlet Dame, fundraising plan and sales delivery insight, Mar 2026) ↩︎
  76. Data moat drives conversion. Boulder conviction. (Scarlet Dame, Martin Kess discovery call, Feb 2026) ↩︎
  77. Organizational context compounds and only works through Aswritten's tools. Boulder conviction. (Scarlet Dame, Martin Kess discovery call, Feb 2026) ↩︎
  78. Introspect, annotate, compile, and story generation form the tool hierarchy. Boulder conviction. (Scarlet Dame, context architecture sessions, Feb 2026) ↩︎
  79. Foundation conviction. (Scarlet Dame, fundraising plan and sales delivery insight, Mar 2026) ↩︎
  80. Foundation conviction. (Scarlet Dame, fundraising plan and sales delivery insight, Mar 2026) ↩︎
  81. Foundation conviction. (Scarlet Dame, fundraising plan and sales delivery insight, Mar 2026) ↩︎
  82. 3-5 weekly calls as primary GTM motion. Boulder conviction. (Scarlet Dame, consultant-led sales and pricing, Feb 2026) ↩︎
  83. Foundation conviction. (Scarlet Dame, fundraising plan and sales delivery insight, Mar 2026) ↩︎
  84. Beta cohort came through advisor introductions, founder networks, existing relationships. Foundation conviction. (Scarlet Dame, Daniel Donoghue beta onboarding, Feb 2026) ↩︎
  85. Direct network is primary channel through F&F raise period. Foundation conviction. (Scarlet Dame, fundraising breakpoints, Feb 2026) ↩︎
  86. Enterprise pipeline through Tony Maley's relationships starts with $10K extraction setup. Stake conviction. (Scarlet Dame, weekly session and Verity demo, Mar 2026) ↩︎
  87. Stake conviction. (Scarlet Dame, weekly session and Verity demo, Mar 2026) ↩︎
  88. Tony's channel activates as involvement formalizes. Boulder conviction. (Scarlet Dame, consultant-led sales and pricing, Feb 2026) ↩︎
  89. Concierge-to-product transition: early customers through high-touch, progressively automating. Foundation conviction. (Scarlet Dame, fundraising plan and sales delivery insight, Mar 2026) ↩︎
  90. Sales motion shifts from founder-led to AI-led as automated intake improves. Foundation conviction. (Scarlet Dame, fundraising plan and sales delivery insight, Mar 2026) ↩︎
  91. Same expansion path grows accounts and grows the company from founder-led to product-led. Foundation conviction. (Scarlet Dame, fundraising plan and sales delivery insight, Mar 2026) ↩︎
  92. SAFE structure confirmed: SIC (C Corp) for product/SAFEs; PnD (S Corp) for consulting income. Every dollar of investment goes to product. Foundation conviction. (Scarlet Dame, fundraising plan, Mar 2026) ↩︎
  93. Raise framing: raw innovation with traction → specific product with sales motion → angel/seed from case studies and unit economics. Boulder conviction. (Scarlet Dame, sales call flow sessions, Mar 2026) ↩︎
  94. "It is not appropriate to take a gift. It IS appropriate to offer a SAFE or investment vehicle where upside is offered based on genuine belief." Boulder conviction. (Scarlet Dame, fundraising reframe sessions, Feb 2026) ↩︎
  95. Demo shows working product, not vision — validates Product NOW focus. Foundation conviction. (Scarlet Dame, YC demo strategy, Feb 2026) ↩︎
  96. The sales motion and product delivery are the same thing. Every sales call builds the prospect's collective memory before they say yes. Foundation conviction. (Scarlet Dame, fundraising plan, Mar 2026) ↩︎
  97. Beta is not self-service onboarding; it is delivering the full consulting engagement so the product is tested under real load with real organizational knowledge. Foundation conviction. (Scarlet Dame, beta strategy reframe, Feb 2026) ↩︎
  98. Onboarding mode is the highest-leverage product gap for beta. Technical infrastructure works; cognitive bootstrapping is what needs iteration. Foundation conviction. (Scarlet Dame, Martin Kess onboarding call, Feb 2026) ↩︎
  99. Outreach to warm leads through advisor network. Revenue-first strategy: consulting engagements validate product-market fit. Boulder conviction. (Scarlet Dame, consultant-led sales and pricing, Feb 2026) ↩︎
  100. "Sell a $5K engagement NOW. Revenue before raise is stronger than projections about future revenue. Investment package becomes stronger with even one paid engagement." Foundation conviction. (Scarlet Dame, fundraising plan, Mar 2026) ↩︎
  101. The quality of the engagement IS the product validation. Foundation conviction. (Scarlet Dame, beta strategy reframe, Feb 2026) ↩︎
  102. $400K funding ladder: senior technical resource converts prototype to scalable infrastructure. Foundation conviction. (Scarlet Dame, fundraising breakpoints, Feb 2026) ↩︎
  103. Pilot pricing at $10K is settled; grounded in rate parity and procurement threshold analysis. Boulder conviction. (Scarlet Dame, revenue model, Feb 2026) ↩︎
  104. Every sales call builds collective memory for the prospect's repo. The free trial is experiencing the product during the sales process itself. Foundation conviction. (Scarlet Dame, fundraising plan, Mar 2026) ↩︎
  105. "Fundraise with a working demo — not a deck about a vision, but a video of a working product." Foundation conviction. (Scarlet Dame, March 2026 priorities) ↩︎
  106. $650K funding ladder: the product is deployed, a real customer is running it, not evaluating it. Boulder conviction. (Scarlet Dame, fundraising breakpoints, Feb 2026) ↩︎
  107. First session is consultant-led; subsequent sessions the user can drive. Automated intake progressively reduces founder dependency. Boulder conviction. (Scarlet Dame, enterprise positioning) ↩︎
  108. Stage-gate logic: each phase produces the evidence for the next raise or sustains on revenue. Foundation conviction. (Scarlet Dame, fundraising plan, Mar 2026) ↩︎
  109. Seed round target no more than $600K, citing effective small-team operation. Stake conviction. (Scarlet Dame, fundraising breakpoints, Feb 2026) ↩︎
  110. SIC for product/SAFEs, PnD for consulting income. Services revenue sustains the founder while product revenue builds the company. Boulder conviction. (Scarlet Dame, fundraising plan, Mar 2026) ↩︎
  111. Solo founder capacity ceiling identified as structural constraint. Boulder conviction. (Scarlet Dame, revenue model, Feb 2026) ↩︎
  112. Solo founder capacity maxes at ~1.5 concurrent engagements; breaking this requires hiring. Boulder conviction. (Scarlet Dame, revenue model, Feb 2026) ↩︎
  113. Breaking into aggressive growth requires Tony or a hire. Boulder conviction. (Scarlet Dame, revenue model, Feb 2026) ↩︎
  114. Tony's prediction: "You've got a two year window" before legal/corporate pressure reshapes the AI landscape. Boulder conviction. (Tony Maley, weekly call, Feb 2026) ↩︎
  115. Settled view based on prior experience with the internet boom pattern. Boulder conviction. (Tony Maley, weekly call, Feb 2026) ↩︎
  116. Pre-revenue confirmed — beta is full consulting engagement, not self-service. Foundation conviction. (Scarlet Dame, beta strategy reframe, Feb 2026) ↩︎
  117. Beta delivers the full $15K consulting engagement at no charge to fully test someone using it with full utility. Foundation conviction. (Scarlet Dame, beta strategy reframe, Feb 2026) ↩︎
  118. Technical debt decision: the worst state is neither committed to current architecture nor migrating. Must be made explicitly. Stake conviction. (Scarlet Dame, roadmap integration sessions, Feb 2026) ↩︎
  119. Scarlet built and refactored the core pipeline solo, including Demo 1 ingestion pipeline in February 2026. Foundation conviction. (Scarlet Dame, pipeline refactor sessions, Feb 2026) ↩︎
  120. Decision between current n8n architecture and Clojure migration must be made explicitly, not by default. Stake conviction. (Scarlet Dame, roadmap integration sessions, Feb 2026) ↩︎
  121. Downside is not zero. Failure floor is a senior AI consultant with proprietary system and case studies. Stake conviction. (Scarlet Dame, fundraising plan, Mar 2026) ↩︎
  122. Stake conviction. (Scarlet Dame, fundraising plan, Mar 2026) ↩︎
  123. Proprietary system, 7 org case studies, deep deployment knowledge cited as durable moat even in failure scenario. Stake conviction. (Scarlet Dame, fundraising plan, Mar 2026) ↩︎
  124. PnD consulting practice is stronger than before SIC existed. Stake conviction. (Scarlet Dame, fundraising plan, Mar 2026) ↩︎
  125. Stake conviction. (Scarlet Dame, fundraising plan, Mar 2026) ↩︎
  126. Subscription pricing is an open question — current tiers need real-buyer validation. Notion conviction. (Scarlet Dame, consultant-led sales and pricing, Feb 2026) ↩︎
  127. Enterprise compliance (GDPR, data sovereignty, ISO) confirmed as real adoption gate by enterprise customer. On-prem deployment option and architecture split are required selling points. Boulder conviction. (Mike Sackton, advisor feedback call, Feb 2026) ↩︎
  128. Boulder conviction. (Mike Sackton, advisor feedback call, Feb 2026) ↩︎
  129. Scarlet manually does interviews; engineer automates pipeline later. The consulting IS the sales motion and the manual delivery IS the product validation. Foundation conviction. (Scarlet Dame, fundraising plan, Mar 2026) ↩︎
  130. Open strategic question: charge early vs. place with high-visibility orgs for free to build leverage and case studies. Foundation conviction. (Scarlet Dame, fundraising breakpoints, Feb 2026) ↩︎
  131. Foundation conviction. (Scarlet Dame, fundraising breakpoints, Feb 2026) ↩︎
  132. Series A operations people have acute pain: fragmented context (Notion hodgepodge), high decision velocity, CEO chaos, already live in ChatGPT. Sits at intersection of customer knowledge and internal process. Stake conviction. (Martin Kess, discovery call, Feb 2026) ↩︎
  133. Billy identified DevOps as valuable persona not currently emphasized; requires validation with DevOps practitioners. Stake conviction. (Billy Sylvester, advisor call, Jan 2026) ↩︎
The link has been copied!