Inspect the evidence before you believe the claim.
Every status, benchmark and control PLLAY makes a claim about — centralized here, cited against the code, so the commercial pages don't have to repeat the caveats.
Sixteen capabilities. One honest matrix.
Available, Managed Pilot, In Development or Research — the one status vocabulary every page on this site cites.
The request/response path a partner integration runs today — open a pool, collect and verify entrants, report the outcome — is live production code, not a design document.
Key-based authentication (Authorization: Bearer pllay_sk_…) is deployed and admin-provisioned per partner.
@pllay/sdk v0.1.0 is built and internally tested, but not published to npm — no partner can npm install it today.
partner-embed renders from b2b_partners.brand_config, with theme governance built on top — but its only b2b_partners/partner_creator_links rows are test fixtures, no real pilot.
capability_flags.pulse_brand_campaigns_enabled = true, but every sponsor_campaigns row in production is demo/pilot data (advertiser is null on all of them); zero campaigns have reached campaign_status = 'COMPLETED' with a signed io_reference and a paid invoice.
Every partner today enters at stage one — a scoped, paid engagement run directly with the PLLAY team. No partner has completed all five stages yet; Campaign OS, the interface stage four and five would run through, is itself in development.
Creator accounts are approved by hand, not automatically. We onboard in small batches, so there is a real wait — you'll hear from us either way.
AWAITING_CREATOR_CONFIRMATION is a real prediction state and settlement is hard-gated on it; authority enforced in platform-api.
ACCURACY_LEDGER: 38 recorded broadcasts, 31 auto-tier, 0 wrong auto-settles. All clips are VODs.
platform_config.vision_autonomous_enabled = false in production; the ledger's 38 trials are all recorded video, none live.
All 38 of Vision AI's benchmark trials are recorded video — a clip is the case the resolver is actually validated on.
The underlying QR / second-screen mechanism is live in production — mobile web, no install, is how every fan on PLLAY plays today.
Winners are paid, the revenue split applies per partner configuration, and the ledger is written — live on every settlement today.
pulse_brand_campaigns_enabled is true for an internal lighthouse pilot only (pilot_entity = 'demo_placeholder'); sponsor-portal exists with no public entry point. Zero sponsor_campaigns rows have reached 'COMPLETED' with a paid invoice.
The permission layer this needs is built and switched off. What is missing is the part that proposes — no agent has ever drafted a question on PLLAY.
No optimization or recommendation capability exists, built or scoped, in this codebase today. The measurement layer it would need is real (see Platform Intelligence, in development) — Creator analytics are live in creator-platform, and the partner-facing surface (GET /v1/intelligence/summary, moments, audience, POST /v1/intelligence/exports) is deployed to production, gated behind scoped API keys and per-partner commercial terms (partner_commercial_plans).
The bench record, condensed — full detail lives at /validation.
Ten platform modules. One system loop.
Full breakdown — integration patterns, enterprise controls, performance targets — lives at /platform.
Authentication, rate limits, webhook delivery and tenant routing for every partner request.
Validate, normalize and route the events a partner's system reports.
Convert structured signals, clip markers and authorized video evidence into a campaign-ready moment.
Manage targeting, schedules, placements, approvals, creative and budget for a sponsored campaign.
Control eligibility, sessions, responses, lock timing and state transitions for a live interaction.
Apply outcome rules, preserve the evidence behind a settlement, and hold or refund when a call isn't clear.
Reserve budget, issue rewards and write the settlement ledger.
Detect abuse, flag suspicious activity and protect campaign integrity.
Measure exposure, participation, outcomes, rewards, cohorts and conversion.
Give an operator one interface for campaign setup, monitoring and reporting.
- Event→
- Eligibility→
- Interaction→
- Response→
- Lock→
- Resolution→
- Reward→
- Measurement
How a design partner engagement actually runs.
Named after Paid design partner, the real stage one of the deployment model.
A first commercial engagement, scoped and paid, with a named integration goal and success criteria set upfront with the PLLAY team.
PLLAY provisions a partner-specific, non-production integration environment and API credentials directly with our engineering team — there is no public self-serve sandbox.
The integration runs against real content, with a defined audience and duration, measured against the objective set in stage one. This is the stage PLLAY's Proof Sprint runs today. Every partner today enters as its own track, not a batched cohort — Every partner today enters at stage one — a scoped, paid engagement run directly with the PLLAY team. No partner has completed all five stages yet; Campaign OS, the interface stage four and five would run through, is itself in development.
Objectives and success criteria are set with the PLLAY team at scoping (stage one), not standardized publicly — there is no self-serve benchmark to compare against.
Results are reviewed directly with the PLLAY team. Sold as a managed enterprise engagement — there is no self-serve signup, and no partner has completed a real production export as of this writing.
The pilot's scope becomes the production integration — the same architecture and credential pattern, promoted rather than rebuilt.
Seven controls, each pointing at a real mechanism.
Plain-language hub at /trust; mechanism-level detail at /security.
A published privacy policy, consent controls and a do-not-sell path ship today across every PLLAY surface.
Settlement writes a ledger entry on every pool; Vision AI's accuracy ledger records all 38 benchmark trials. This is an internal record today, not a published, partner-facing audit-log export.
A reward-budget field exists in the sponsor portal's source, but that portal isn't deployed — reward budgets are set per campaign directly with the PLLAY team today.
Fraud and anomaly detection runs on a recurring schedule in production, flagging suspicious activity for review before it reaches a payout.
Capability flags already gate major functions in production — vision_autonomous_enabled, pulse_brand_campaigns_enabled — but there is no partner-facing pause control; today a flag change is an internal, PLLAY-operated action.
Settlement writes a ledger entry on every pool; Vision AI's accuracy ledger records all recorded-broadcast trials — see Vision AI validation above.
The schema is multi-tenant by design — per-partner credentials and brand configuration — but it has never been exercised by more than one real external partner at a time, so isolation at production scale is unverified.
No published external production case studies yet.
capability_flags.pulse_brand_campaigns_enabled = true, but every sponsor_campaigns row in production is demo/pilot data (advertiser is null on all of them); zero campaigns have reached campaign_status = 'COMPLETED' with a signed io_reference and a paid invoice.
This section is built and ready — the first entry ships the day a real one closes.
Become a Design Partner
A first commercial engagement, scoped and paid, with a named integration goal and success criteria set upfront with the PLLAY team.