Post → Support → Escalation
Ideas gain visibility as they earn support. Thresholds adjust with context so the best contributions rise without flooding the feed.
COGENSY turns public deliberation into a tamper evident, machine readable feed that preserves both consensus and principled dissent. Simple loop: post → support → escalation, with bot friction, reputation (time decay), and diversity of support thresholds to resist brigading. Outputs ship as weekly/final briefs (PDF) and a read only JSON API.
COGENSY is a human centric consensus layer. It captures, filters, and promotes ideas based on breadth and quality of support, not raw volume. Every promotion is cryptographically fingerprinted and published, so anyone can verify the record of how a statement rose to consensus—without exposing identities. Institutions and AI systems read a simple consensus + dissent digest via PDF/JSON to inform policy and alignment workflows.
Ideas gain visibility as they earn support. Thresholds adjust with context so the best contributions rise without flooding the feed.
Promotion requires backing from distinct cohorts (e.g., stakeholders, regions, time buckets) to reduce brigades and astroturf.
Constructive patterns earn influence that decays over time; bursty or low value behavior loses weight.
Promotions are hashed and batched into published Merkle root fingerprints; a verifier page recomputes and confirms integrity.
Participants post ideas or support others. Bot friction and a moderation gateway filter obvious abuse before it hits the feed.
Items promote only when they meet reputation weighted and diversity of support thresholds, with stability checks across snapshots.
Every promotion is hashed; batch Merkle roots are published on a cadence. Anyone can verify a promoted item’s fingerprint.
Institutions receive weekly/final briefs (PDF) and can subscribe to a read only JSON API for dashboards, evaluation, or RLAIF pipelines.
No PII required to participate. Minimal logs. Data residency can be constrained per partner region.
Thresholds, decay curves, and moderation outcomes are inspectable; dissent is preserved alongside consensus.
Outputs include: 5–10 consensus statements each with linked dissent, diversity scores, stability checks, and a verifier link.
Offered under Reg D 506(c) to accredited investors only. This is not an offer to sell securities in any jurisdiction where prohibited. By continuing, you attest you are an accredited investor.
Run a 30 day, no cost pilot on a live policy topic (e.g., mental health, housing). One line embed on your site; weekly digests and a final consensus+dissent brief with public fingerprints.
Read only API ingestion of consensus+dissent digests from a live civic topic. Use for eval comparisons, adversarial prompts (dissent), and RLAIF calibration. No PII, no cost, no exclusivity.
Thank you for confirming accreditation. Below are materials for your review. Please note that all offerings are made only pursuant to final documents.
Read the canonical overview and MVP plan.
Standard post money SAFE (506(c)); wire/ACH handled by our investment portal.
Questions or diligence requests.
One live topic on your site via a one line embed. Participants post and support; items escalate when thresholds are met. We deliver two weekly digests and a final brief (PDF + JSON), with public fingerprints and a verifier link.
5–10 consensus statements with linked dissent; diversity of support met across cohorts; stability across snapshots; public audit proof included in your “What we heard” or staff report.
We provide copy for your page and the embed; you approve the prompt and moderation rubric.
30 day read only API access to consensus+dissent digests from a live civic topic; weekly JSON + final pack. Use dissent as adversarial prompts; use diversity weighted consensus for evaluation or RLAIF calibration.
No PII; rate limited keys; no publicity or logo use without written consent; non exclusive; zero cost.
We can share a schema and a short LOI template to kick off.