For investors
The proof layer for human conversations.
AI made transcripts of meetings that never happened free to produce. Every reviewer of qualitative research, journalism, and student work now needs a cheap way to tell real fieldwork from fabricated fieldwork. We sell that check, and we sell it one layer below the apps that will be copied.
Pre-launch and pre-revenue. Version 0.2 is built; nothing below is a forecast.
The problem
Research-integrity tools verify documents. None verify the meeting.
Hashes, timestamps, and registries prove a file existed. They cannot prove that distinct real people were present for a span of time and endorsed a specific version of the record. Interviewees, meanwhile, have no way to say “I was there and that is what I said” without giving up anonymity. That gap is the product.
The product in three layers
Personhood, attestation, application.
Cubid, existing
Proof of personhood, sybil resistance, humanity score, pairwise app-scoped identifiers. Every participant is a Cubid user. Cubid is a separate, existing company.
ProofOfMeetProtocol, infrastructure
Workspaces, Merkle-committed records, an append-only commit chain, signed attestations with witness statements, disclosure profiles, daily anchoring into Bitcoin. A small UI; the product is the API, SDK, and MCP server.
VerifyMyResearch, application
Studies that aggregate meetings into a methodology disclosure, with research-shaped roles from researcher to ethics board to public. Integrator number one of the infrastructure, and the template for every copy that follows.
The moat is the seam, not the app
Anyone can build a meeting-verification app in a weekend. We accept that.
The defensible value sits one layer down. The server is private; the SDK, API client, MCP server, and protocol spec are Apache-2.0. A competitor to VerifyMyResearch is a customer of ProofOfMeetProtocol. Document-verification platforms are integration targets, not rivals, and plugging in is a one-day job. Even a clone of the whole stack still needs proof of personhood, which keeps Cubid in the loop and keeps the standard ours.
- Open at the edges, private at the core
- VMR source-available after a first-mover period, on a published schedule
- Every copy drives volume to the attestation layer and to Cubid
- Federated sign-in across apps removes the second-account problem
Business model
Participants never pay. Institutions are the core line.
Designed in the whitepaper, not yet enforced; pre-launch versions take no money.
| Who | Pays | Gets |
|---|---|---|
| Participant | Nothing, ever | Attest, set disclosure, view own meetings |
| Individual researcher or journalist | Roughly 10 to 20 USD per month | VerifyMyResearch study plus ProofOfMeetProtocol workspaces |
| Institution | Annual seat-based license | All students or staff, admin console, SSO, data residency, audit roles |
| Integrator | Metered API with a free developer tier, or revenue share | API, SDK, and MCP at volume |
Public proofs
Anchored in Bitcoin, the most secure blockchain there is.
Every daily batch root is signed with our published key and submitted to OpenTimestamps, which aggregates it into a Bitcoin transaction. No token, no wallet, no fee per proof, and no dependency on us staying online: once a root is in a Bitcoin block, anyone can verify a receipt with standard open tooling for as long as Bitcoin exists. Pending proofs are stored today; the automatic upgrade to confirmed proofs ships in a pre-launch version.
- Bitcoin through OpenTimestamps; no token or wallet involved
- One transaction carries any number of proofs, so cost does not grow with volume
- Verification needs only the open SDK and the Bitcoin chain
Roadmap
Six phases. We are inside the first.
The whitepaper roadmap, with each phase marked as it stands today. Phase 0 is built; phase 1 is where the work is.
- 0. FoundationsRepos, Cubid sign-in, data model, Merkle commitments, commit chain, minimal UI, REST API, batch anchoring to Bitcoin via OpenTimestampsBuilt
- 1. ProofOfMeetProtocol launchPublic verification pages, disclosure profiles, SDK and MCP server, embeddable component, individual subscriptionsIn progress
- 2. VerifyMyResearch launchStudies, research roles, notetaker ingestion, methodology reports, federated sign-in, bundled pricingDesigned, not built
- 3. InstitutionsEnterprise licenses, admin console, SSO, data residency, audit and ethics-board roles, consent exportDesigned, not built
- 4. Stronger proofsParticipant-held keys, zero-knowledge score thresholds, confirmed Bitcoin proof upgrades, independent anchoring monitorsDesigned, not built
- 5. Growth linesScheduling, paid public API at scale, native recording and transcription, VerifyMyResearch source releaseNot yet
Where we are today
What has been built, what has not, and what we do not know.
Version 0.1 and 0.2
Built in November 2026: the full attestation core, commit chain, provenance, grades, API, SDK 0.2, MCP server, protocol spec 1.1, test vectors, and local end-to-end tests.
Deployment and partners
First hosted deployment, production Cubid sign-in, and one or two university departments as design partners. Owners are allowlisted; the first-ninety-day target is twenty owners and a majority of invited participants attesting.
Revenue, users, launch date
No revenue, no public users, no launch date. Pricing is designed and not enforced. The founder is deciding which pre-launch version becomes the launch.
Open questions
Immutable proofs versus the right to erasure under privacy law; collusion between an owner and real but complicit participants; how much of the proof moves to participant-held keys and zero knowledge.
Express interest.
We are talking to a small number of people about a first round, partnerships with notetakers and research-integrity platforms, and advisory roles in research ethics and journalism. Tell us which, and we will reply personally.