This document was written with Claude reading from Aswritten's own perspective: the extracted, attributed record of the company's decisions and reasoning. Every claim below carries a citation back to a specific person, context, and conviction level. The citations are not decoration. They are the product, run against its maker.

The Opportunity

The consensus is settled: AGI is coming, and there will be an AI for nearly every task. The real question is not whether we reach superintelligence. It is how many voices are inside it. Today a handful of frontier labs produce the models that mediate our interactions, and every one of those models answers from the middle of its training data. Your organization does not live in the middle. Neither do you.[1]

Aswritten's contrarian bet: the number of voices can be as large as the number of people and organizations willing to encode their own. You build this on top of the monoculture rather than against it. Treat the model as hardware and a perspective as the program you install on it. The same mechanism that lets an enterprise install its actual judgment onto AI lets a person install theirs. The moonshot is not an AI for every task. It is an AI for every task that is as different as we are different.[2]

Every organization deploying AI is already building its own approach to context: knowledge bases, project instructions, documentation, integrations. Those tools can only reach what someone wrote down, and most of what an organization knows was never written down. The reasoning behind a decision, the methodology, the "why we did it this way," stays in the heads of the people who hold it. When they are busy, everyone waits. When they leave, the knowledge leaves with them.[3]

OpenAI validated the category in early 2026 when it invested in Frontier to build a shared context layer for enterprise AI agents. Frontier assumes the knowledge already exists in enterprise systems. It does not. Someone has to extract it first. That extraction is the opportunity, and it sits upstream of every agent platform.[4]

What I Built

Aswritten turns an organization's conversations into a perspective: its decisions and reasoning, written down with attribution. Who said what, when, in their own words. Through structured conversational capture, the system draws out undocumented knowledge, then compiles it into a versioned, queryable perspective that any AI tool across the organization can read through the Model Context Protocol.[5]

Every claim in the perspective carries a conviction level and full provenance. A passing thought is marked differently from a settled architectural decision. The scale runs from notion (easily moved, a first mention) through claim (asserted, still validating) and decision (settled, load-bearing) to principle (bedrock). Every claim traces to the person who made it, the context they made it in, and what it replaced. This is not a wiki that goes stale. It is a living record where knowledge has weight, attribution, and history.[6]

When an AI answers from the perspective, it does more than know more facts. It reasons with the organization's actual judgment. The cite tool answers the question of where a claim came from: click the citation, read the original words from the person who said them. Checking the AI stops being a matter of trusting the model and becomes a matter of trusting the person it quotes. The introspect tool answers the inverse: it names the gaps the perspective has not filled yet, so the AI knows the edge of what the organization has settled.[7]

The perspective lives in the customer's own repository. The AI runs on the customer's own side of the wire: their local Claude, their enterprise OpenAI. Aswritten holds nothing at rest. The perspective is rebuilt in memory from the customer's own transactions, served, and released. On-prem, air-gapped, and sovereign deployments are available for regulated and high-trust engagements, so transcripts and AI history never leave the customer's server.[8]

The perspective is also versioned and branchable, the way developers branch code. A team can propose a shift, review how it moves the worldview, and build consensus before merging. The product is live, in production, and works across multiple AI platforms rather than locking to any single vendor. This document was written from Aswritten's own perspective.

Why Me

I started programming the flocking of birds in 2011 and watched simple local rules, do not collide, do not stray too far, align with your neighbors, produce complex and beautiful behavior no one could predict from the top down.[9]

I spent years working with language as code. With LLMs in 2019 (then GPT-2) I saw that these machines built to predict the next word carried a kind of velocity, steered by narrative: the directional vector of meaning beneath statements that mean far more than their single sentence.[10]

Aswritten is my fifth 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. I taught myself to program in the third grade and spent my first professional decade writing software in data visualization, natural language processing, and novel user experience across existing technology. Since 2015 I have mixed software consulting with entrepreneurship, and in 2022 I formalized a narrative strategy consultancy.[11]

As Chief Strategist at Vouch.io, I developed a manual process for steering LLMs with narrative. Instead of training new models, I treated the model like a computer and the narrative architecture like a program: an interconnected web of narratives that defined a way of thinking. I used it to bootstrap any AI to think like the company instantly, to create content, test ideas, even write code. Instead of defaulting to the average of the internet, the model was now flying in the direction of the company.[12]

I left Vouch in September 2025 to automate that process, and built Aswritten. Every conversation with a coworker, every piece of client feedback, every strategic argument, every engineering decision is a course correction that currently evaporates. Aswritten catches the heading change before it is lost and turns it into a perspective that points toward where the organization is going, past where it merely is today.[13]

I know the product works because I use it every day. I built the platform solo. AI gave me speed, and Aswritten gave AI a controllable direction. It is what I needed for my team at Vouch, and it is what I hear from advisors, pilot customers, and sales conversations as I move toward product-market fit.[14]

LLMs feel like flying. But they moved our starting line from the blank page to the complete and misguided draft. I built Aswritten to feather our arrows, so the first attempt flies true. Our direction, as written.

Traction & Proof

Aswritten is no longer pre-revenue. On April 30, 2026, an external customer paid for the first time. A production pilot with a paying enterprise customer converted to a paid subscription within a month, with an organization-wide contract staged behind it.[15]

The pilot runs live inside the customer's architecture review process. Engineers reach the perspective through GitHub Copilot; product owners reach it through Claude desktop. The conversations that already steer that organization now also steer its AI, and the AI's answers carry citations back to the people who made the calls. This is the product running under real load with real organizational knowledge.[16]

The pilot proves the motion. Every capture the customer commits, every decision they trace, every teammate who queries the perspective, all of it builds context that only works through Aswritten's tools. The customer is not evaluating a demo. They are running a product that already knows their business, and the accumulated perspective is what pulls them from pilot to organization-wide contract.[17]

The category has independent validation. OpenAI's Frontier investment confirmed shared organizational context for AI agents as an enterprise requirement. Aswritten sits upstream of it, solving the extraction problem the agent platforms do not touch.[18]

The pilot also confirms a structural advantage in how Aswritten sells. The sales call, the demo, and the delivery are the same activity. A discovery engagement gathers senior knowledge and extracts it into a working perspective inside the conversation itself. The pipeline is also product validation.[19]

Business Model

Aswritten's model rests on a structural insight: the work that onboards a customer is the same work that delivers the product's core value. A consultant-led session gathers the raw material, interviews with senior staff, existing documents, decision history. The product handles what happens next, extracting that material into a conviction-weighted, provenanced perspective. That engine is built and running. The gathering fills the tank; it does not build the engine.[20]

By the end of an engagement, the customer has a working perspective, and every AI tool in the organization can query it. There is no gap between sales and delivery. The setup work is the value.[21]

The product sells across tiers that map to a natural expansion path, from a single private user, to organization-wide deployment with a knowledge extraction package, to enterprise deployment with no user limits and requirements-driven options for on-prem, air-gapped, and sovereign installs. Customers control token spend by bringing their own AI vendor key through OpenRouter, which keeps margins predictable and gives enterprise customers direct control over their own usage.[22]

The expansion motion is organic. A user discovers Aswritten and connects their own key. They start a private perspective, invite a teammate, and grow into an organization tier. As the perspective compounds across the team, they want bespoke automations and cross-team integration, and they move to enterprise. Each step up is driven by the perspective the customer has already built, which becomes a data moat that only Aswritten's tools unlock.[23]

The enterprise product carries the company. The same research-and-development line serves a second, longer-horizon product: an individually owned perspective, a companion a person owns, evolves with, and can share. R&D line equals product line. Nothing built for the enterprise extent is thrown away when it serves the individual one. The organizational case simply adds the machinery of audience, clearance, and role. That is why the moonshot and the revenue are a single bet.[24]

Go-To-Market

The three enterprise use cases are one perspective under three verbs, in sequence.[25]

Discovery creates the perspective. It is what a consultancy already does: go into an organization, run interviews, produce transcripts, and synthesize them into a plan quoted back to source. Aswritten automates that synthesis to high fidelity and lets people use AI where it would otherwise be a hazard, because every claim traces back to source instead of hallucinating past it. Discovery is the entry engagement.[26]

Onboarding transfers the perspective. Once the knowledge base exists, an organization can onboard new staff and new AIs into its perspective. New humans and new AIs are the same recipient shape. Onboarding is the expansion.[27]

Governance maintains and steers with the perspective. This is the architecture-review motion the pilot runs today: the organization is already using conversations between humans to steer itself, and now those same conversations steer its AIs, which handle work from internal QA to answering the team's questions about a decision. Governance is the subscription.[28]

Read as a sales arc: discovery enters, onboarding expands, governance subscribes. It is a single-customer sequence rather than three separate segments. Two channels feed it. The direct network, built from advisor introductions, founder relationships, and existing trust, converts warm conversations into organization subscriptions on a short cycle, because the call itself is the proof. The enterprise pipeline, opened through advisor relationships and industry networks, starts with a discovery engagement and expands into cross-team integration and deployment options: a longer cycle, higher contract value.[29]

As the automated intake improves, refined through real usage and the patterns drawn from human-led session transcripts, the motion shifts from "get on a call with Scarlet" to "connect your repo and start talking to the AI." The same expansion path that grows an account from a single user to enterprise also grows the company from founder-led sales toward product-led growth.[30]

The Ask

I am raising a $500K target, structured as a SAFE plus a shared-earnings instrument, on an $8 to 12M cap with no discount, with checks starting at $25K. The raise is explicitly incremental: any check counts, and any single significant win substantially extends the runway.[31]

The bridge arrives holding two proofs at once: purchased enterprise traction on the current architecture, and a solo-partner proof-of-concept of the next architecture, built now, pre-raise, on no new money. The round funds applying that proven architecture across the full enterprise extent, clearance, federation, role topology, cross-repo, and ships it to the paying customer as upgrades as each piece proves out. This upgrades existing revenue along a proven path rather than opening a second bet.[32]

This is not a gift. It is an investment in a working product with a paying customer, offered on standard terms with standard protections. The instrument embodies the thesis: aligned capital, optionality preserved, no treadmill. The company chooses when to take money next, because the calibration discipline makes it fundable precisely by not needing to be funded.[33]

Use of Funds & Phases

The bridge funds a small, senior, part-time team, all aligned through profit-share so cash burn stays calibrated to validated demand. Part-time enterprise-software help and part-time enterprise delivery absorb the client-driven work, which frees me to build core product and sell. Team sizing and pricing are calibrated to demand that has actually been purchased, never to a venture timeline: no bloating ahead of the pipeline, no team growth before customers justify it.[34]

Now through summer. Convert the pilot's organization-wide contract, land a second pilot on the same motion to prove repeatability beyond one customer, and stand up a consulting-discovery wedge: bounded engagements with faster cycles that productize existing consulting strength and seed future organization-tier customers. The bespoke analytics delivered during the first customer's contract review double as the discovery deliverable's live prototype and as case-study material.[35]

Fall. Pilots and team-tier conversions produce the evidence a larger round requires: named customers, retention, expansion revenue, repeatable motion. The automated intake keeps improving, reducing founder dependency on every onboarding. A reapplication to Y Combinator lands on this same timeline, deliberately: the prior application drew a top-decile rejection with "come back with traction," and the traction now exists.[36]

Into the new year. Warm the layer for an institutional round of $3 to 4M, sized to carry the full team across a multi-year window, and close it from strength rather than survival. If the venture path is not ready, calibrated consulting and subscription revenue sustain the company while the perspective keeps compounding. Either path is forward.[37]

Risk & Downside

The honest risks:

Customer concentration. The first paying customer anchors several paths at once, so a churn there is correlated risk across paths rather than an isolated event. The mitigation is structural: a second pilot and the consulting wedge diversify revenue, multiple stakeholders inside the account stay engaged, and the motion is documented in transferable form rather than living only in my head.[38]

Solo founder. Every path routes through one person, and variable pace is the operating reality. The parallel-path posture holds slack by design, handoff docs let work advance without me, and senior part-time hires arrive at bridge close. The structure has to survive slow weeks; that is a design requirement built into how the company runs.[39]

Repeatability unproven. One customer does not establish repeatability, and an institutional round likely raises on partial evidence. A paid engagement sourced outside the immediate network is the cheapest honest repeatability signal available, and the consulting wedge produces exactly that.[40]

Market timing. An industry advisor reads a roughly two-year window before legal and corporate pressure reshapes the AI landscape, pattern-matched to the early internet boom. Moving fast matters more than moving perfectly.[41]

The downside is bounded. In the worst case, the product does not find broader fit and the company winds down. SAFE investors receive standard dissolution protection, and what remains is a senior AI consultant with a proprietary system, real deployment knowledge, and a consulting practice stronger than before the company existed. The floor is a better consulting business than the one that existed before, well short of starting over.[42]

How to Read the Citations in This Document

This document is a product demonstration.

Every footnote traces to a specific memory in Aswritten's git repository. Those memories compile into a perspective: a structured record of what the company knows, believes, and has decided. Each citation names its source (who said it, in what role), its conviction level on the current scale (notion, claim, decision, principle), and its provenance in the graph.

The document was written against the perspective and re-verified with the cite tool: each substantive claim above was checked back to its source memory before it was footnoted. Because the perspective is the source, the document regenerates when the company changes. When a new customer lands, a price shifts, or a strategy turns, the company commits a new memory, and this brief is rewritten and re-verified against it. The diff between one version and the next is a literal record of the company's evolution.

  1. Scarlet Dame, morning-pages voice memo, July 1, 2026: the moonshot reframe. The consensus (AGI is coming, an AI for every task) is conceded deliberately; the contrarian turn is "how many voices in superintelligence are there going to be?" AI answers from "a singular opinion in the middle of its training set," and organizations live nowhere near the middle. Scarlet Dame, moonshot / representation thesis, Jul 2026. decision. ↩︎
  2. Scarlet Dame, same voice memo: "we can build this on top of the monoculture... treating the model like hardware that we install the program of perspective on top of," and the thesis line, "an AI for every task that is as different as we are different." Model as hardware, perspective as program. Scarlet Dame, moonshot / model-as-hardware framing, Jul 2026. decision. ↩︎
  3. The distinction between documentation and a perspective: documentation captures static artifacts; a perspective captures decisions, reasoning, and their rationale as a living record. The highest-value knowledge is undocumented and lives in senior experts' heads. Scarlet Dame, positioning sessions. decision. ↩︎
  4. OpenAI's Frontier investment validated shared organizational context for AI agents as an enterprise requirement. Aswritten's differentiation is upstream: it solves the extraction problem the agent platforms assume away. Scarlet Dame, Frontier competitive analysis. claim. ↩︎
  5. A perspective is the organization's decisions and reasoning written down with attribution: who said what, when, in their own words. The AI reaches it through the Model Context Protocol, grounding the tools the organization already uses. Scarlet Dame, positioning and integration sessions. principle. ↩︎
  6. Conviction levels track how settled knowledge is, on a four-point scale: notion (easily moved), claim (asserted, still validating), decision (settled, load-bearing), principle (bedrock). Every claim carries the person, the context, and what it replaced. Scarlet Dame, architecture sessions; conviction-scale rename. decision. ↩︎
  7. The moment of value is when someone sees AI-generated text with citations showing where it came from. The cite tool ("annotate" renamed to "cite" on Mar 30, 2026, because "cite is what the user wants: verify where claims come from") answers where a claim came from; introspect names the gaps the perspective has not filled yet. Scarlet Dame, provenance/citation moment, Mar 2026; tool-naming session, Mar 2026. principle. ↩︎
  8. Data sovereignty as trust differentiator: the perspective lives in the customer's own repository, the AI runs on the customer's side of the wire, and Aswritten rebuilds the perspective in memory from the customer's transactions, serves it, and holds nothing at rest. On-prem and sovereign deployment is a requirements-driven option, triggered by compliance needs (HIPAA, GDPR), validated by the pilot customer's CTO as a hard requirement for international clients. Scarlet Dame, data-sovereignty and on-prem sessions; customer CTO demo, Apr 2026. principle. ↩︎
  9. Flocking origin (2011): simple local rules producing emergent complexity, priming an understanding of movement in ecosystems and of concepts within frameworks. Scarlet Dame, founding-story through-line, Feb 2026. principle. ↩︎
  10. Core technology insight: language models have velocity, and narrative is the directional vector of meaning beneath statements. Scarlet Dame, org architecture decisions, Feb 2026. principle. ↩︎
  11. Career arc: digital agency, venture-funded acquihire, arts collective, narrative-strategy training program; self-taught programmer; data visualization and NLP; consulting plus entrepreneurship since 2015; narrative-strategy consultancy formalized 2022. Scarlet Dame, founding-story through-line, Feb 2026. principle. ↩︎
  12. At Vouch, Scarlet created a narrative source of truth by hand: an interconnected web of narratives treated as a program installed onto model hardware, so AI defaulted to organizational narratives instead of training-set ones. Scarlet Dame, founding-story through-line, Feb 2026. principle. ↩︎
  13. Aswritten created September 2025 as the automation of the Vouch process: an ingestion pipeline that extracts narrative information from arbitrary content and builds a perspective pointing toward where the organization is going. Scarlet Dame, founding-story through-line, Feb 2026. principle. ↩︎
  14. Founder built the platform solo; validated by feedback from an industry advisor, an agent-systems builder, and the pilot customer's chief architect. Scarlet Dame, pilot plan draft and advisor feedback, Feb–Mar 2026. decision. ↩︎
  15. First paid transaction by an external customer on April 30, 2026: a production pilot with a paying enterprise customer converted to a paid subscription within a month, with an organization-wide contract staged behind it. Scarlet Dame, customer conversion, Apr–May 2026. decision. ↩︎
  16. The pilot runs inside the customer's architecture review process; engineers reach the perspective via GitHub Copilot, product owners via Claude desktop. The system's ability to show how a perspective moves across transcripts was identified by the customer's chief architect as a key product moment. Scarlet Dame, pilot deployment; customer chief architect, pilot kickoff, May 2026. decision. ↩︎
  17. Every capture builds context that only works through Aswritten's tools; the accumulated perspective is the data moat that drives conversion from pilot to organization-wide contract. Scarlet Dame, data-moat and sales-delivery insight, Mar 2026. decision. ↩︎
  18. Shared organizational context for AI agents validated as an enterprise requirement by OpenAI's Frontier investment; Aswritten solves the upstream extraction problem agent platforms do not address. Scarlet Dame, Frontier competitive analysis, Feb 2026. claim. ↩︎
  19. The sales motion and the product delivery are the same activity; every discovery engagement builds the prospect's perspective before they decide to buy, which makes the pipeline product validation. Scarlet Dame, sales-as-delivery insight, Mar 2026. decision. ↩︎
  20. The onboarding work fills the tank, not builds the engine: the extraction engine that turns raw material into a queryable perspective already works; what is still manual is the gathering. Scarlet Dame, onboarding-as-gap sessions, Feb 2026. decision. ↩︎
  21. By engagement's end the customer has a working perspective queryable by every AI tool in the organization; there is no gap between sales and delivery, and the setup work is the value. Scarlet Dame, sales-delivery insight, Mar 2026. decision. ↩︎
  22. Customers control token spend by bringing their own AI vendor key through OpenRouter, reframed from limitation to enterprise selling point; on-prem is a requirements-driven option, not an org-size gate. Scarlet Dame, V1 scope and pricing sessions, Feb–Mar 2026. claim. ↩︎
  23. The expansion track (single user → organization tier → enterprise) is driven at each step by the data moat the customer has already built. Scarlet Dame, expansion-motion and data-moat sessions, Feb 2026. decision. ↩︎
  24. Enterprise revenue carries the company; the same R&D line serves an individually owned product (a companion a person owns, evolves with, and can share). "R&D line equals product line": nothing built for the enterprise extent is discarded for the individual one; the org case adds audience machinery. Scarlet Dame, moonshot / R&D-line-equals-product-line framing, Jul 2026; bridge pitch, Jun 2026. decision. ↩︎
  25. The three enterprise use cases are one perspective under three verbs in sequence: discovery creates it, onboarding transfers it, governance maintains and steers with it. Read as a sales arc, discovery enters, onboarding expands, governance subscribes. Scarlet Dame, use-case taxonomy, Jul 2026. claim. ↩︎
  26. Discovery: automating what a consultancy does (interviews, transcripts, synthesis quoted back to source) to high fidelity, so AI stops being a hazard that produces untraceable hallucinations. Scarlet Dame, discovery use case (call with a consultant who ran discovery engagements for fifteen years), Jun–Jul 2026. claim. ↩︎
  27. Onboarding: enabling an organization to onboard new staff and new AIs into an existing perspective; new humans and new AIs are the same recipient shape. Scarlet Dame, onboarding use case, Jul 2026. claim. ↩︎
  28. Governance: the architecture-review motion, where conversations that already steer the organization now also steer its AIs, which handle work from internal QA to team questions about decisions. This is the pilot customer's use case. Scarlet Dame, governance use case / pilot ARB motion, Jul 2026. claim. ↩︎
  29. Two channels: the direct network (advisor introductions, founder relationships, existing trust) converting on a short cycle because the call is the proof, and the enterprise pipeline (advisor and industry relationships) starting with a discovery engagement and expanding into deployment options. Scarlet Dame, go-to-market channels, Feb–Mar 2026. decision. ↩︎
  30. As automated intake improves through real usage and human-led-session transcripts, the motion shifts from founder-led to product-led; the same expansion path grows both the account and the company. Scarlet Dame, concierge-to-product transition, Mar 2026. decision. ↩︎
  31. Friends-and-family bridge, $500K target, explicitly incremental; SAFE plus shared-earnings instrument; $8–12M cap, no discount; checks from $25K. Any check counts; any single significant win substantially extends runway. Scarlet Dame, bridge structure, Jun 2026. decision. ↩︎
  32. The bridge arrives holding both purchased enterprise traction on the current architecture and a solo-partner proof-of-concept of the next architecture, built pre-raise on no new money; the round applies the proven architecture across the full enterprise extent and ships to the paying customer as upgrades. An upgrade path for existing revenue, not a second bet. Scarlet Dame, bridge pitch, Jun 2026. decision. ↩︎
  33. Investment, not gift: it is appropriate to offer a SAFE or investment vehicle where upside is offered based on genuine belief. The instrument embodies aligned capital, preserved optionality, and no treadmill; the calibration discipline makes the company fundable by not needing to be funded. Scarlet Dame, fundraising reframe and calibrated-growth sessions, Feb–Jun 2026. decision. ↩︎
  34. Calibrated growth: team sizing and pricing calibrated to validated, purchased demand, not venture timelines; profitability at each configuration is the math check, not the destination; the shared-earnings structure and profit-share retainers keep alignment scaling with risk. Part-time senior help absorbs client-driven work to free the founder for core product and sales. Scarlet Dame, calibrated-growth staircase, Jun 2026. decision. ↩︎
  35. Near-term sequence: convert the org-wide contract, land a second pilot on the same motion for repeatability, and run a consulting-discovery wedge of bounded engagements that productize consulting strength and seed org-tier customers; the customer's bespoke analytics are the discovery deliverable's prototype and case-study material. Scarlet Dame, customer strategy and consulting wedge, Jun 2026. decision. ↩︎
  36. Fall produces the evidence a larger round requires (named customers, retention, expansion, repeatable motion); a Y Combinator reapplication lands on this timeline deliberately, following a prior top-decile rejection with "come back with traction," which now exists. Scarlet Dame, funding paths and YC timing, Jun 2026. decision. ↩︎
  37. Institutional round of $3–4M, 6–12 months post-bridge, sized for the full team across a multi-year window, raised from strength and closed in the new year; if venture is not ready, calibrated revenue sustains the company. Either path is forward. Scarlet Dame, funding strategy, Jun 2026. decision. ↩︎
  38. Customer concentration is correlated risk across paths; mitigations are a second pilot and the consulting wedge for diversification, multi-stakeholder engagement inside the account, and the motion documented in transferable form. Scarlet Dame, pilot-customer contingency runbook, Jun 2026. decision. ↩︎
  39. Founder as single point of failure with variable pace as operating reality; parallel paths hold slack, handoff docs advance work without the founder, and senior part-time hires arrive at bridge close. The structure surviving slow weeks is a design requirement. Scarlet Dame, risk register, Jun 2026. decision. ↩︎
  40. Repeatability is not established by one customer; the institutional round likely raises on partial evidence, and a paid engagement sourced outside the network is the cheapest honest repeatability signal, produced by the consulting wedge. Scarlet Dame, repeatability risk and consulting wedge, Jun 2026. claim. ↩︎
  41. An industry advisor's read: a roughly two-year window before legal and corporate pressure reshapes the AI landscape, pattern-matched to the early internet boom. Moving fast matters more than moving perfectly. Industry advisor (long-time collaborator), weekly call, Feb 2026. decision. ↩︎
  42. Bounded downside: worst case the product does not find broader fit and the company winds down; SAFE investors receive standard dissolution protection, and the floor is a senior AI consultant with a proprietary system, deployment knowledge, and a stronger consulting practice than before the company existed. Scarlet Dame, downside floor, Mar 2026. claim. ↩︎