# tvScientific > tvScientific (branded "tvScientific by Pinterest" since Pinterest completed its acquisition > in February 2026) is a connected-TV advertising platform for performance marketers. It pairs > premium streaming-TV inventory with AI-driven bidding and outcome-based measurement so > advertisers can buy television against measurable goals like sales, app installs and site > traffic. ## Read this first tvScientific does **not** publish a developer program. There is no developer portal, no API reference, no OpenAPI or GraphQL contract, no SDK, no CLI, no MCP server, no agent card and no llms.txt of its own. This file was generated by API Evangelist from what a public, anonymous client can actually reach, so an agent does not have to rediscover it. A live REST API nevertheless exists and is reachable: - **Base URL:** https://api.tvscientific.app - **Named by the provider** as the "Campaign Management API" component on https://status.tvscientific.com - **Version roots are public:** https://api.tvscientific.app/v1/ and https://api.tvscientific.app/v2/ each return a JSON index of every resource collection, with no credentials. - **Everything else is credentialed.** All collections but one return `403 {"detail":"Authentication credentials were not provided."}`. The credential type is not published and no `WWW-Authenticate` challenge is sent, so it cannot be determined from outside. Access is arranged with the tvScientific account team. - **Do not attempt to guess credentials.** There is no self-serve sign-up and no public key issuance. ## API surface (as advertised by the API's own version roots) - v1 — 35 collections: advertisers, advertiser_categories, advertiser_users, alerts, allowed_targeting, audiences, audience_types, bid_strategies, bid_strategy_events, billing_accounts, campaigns, contracts, creatives, creative_lineitems, static_display_creatives, static_display_lineitems, image_assets, invoice_emails, organizations, organization_users, payment_profiles, platform_invites, referral_sources, referral_source_costs, tenants, terms_and_conditions, user_agreements, users, user_types, video_assets, plus four staff-facing `internal/` collections. - v2 — 15 collections: campaigns, campaigns/experiment, lineitems, creatives, creativelineitems, staticcreativelineitems, staticdisplaylineitems, videoassets, audiences, allowed_targeting, targetingkeys, targetingreach, simplifiedbidstrategies, bidder-config, quickstart. - Both major versions are live at the same time. No policy says which one to use. ## Runtime conventions (observed, not documented) - Paths require a trailing slash. - Pagination: `?page=` and `?page_size=`; responses wrap records as `{count, next, previous, results}`. - Records are hyperlinked: each carries an absolute self `url`, and relationships are absolute resource URLs rather than bare ids. - Content negotiation via `Accept` or `?format=json` / `?format=api` (browsable HTML). - Errors are `{"detail": "..."}` — not RFC 9457, no machine-readable error code. Routing 404s return HTML, so a JSON client must tolerate a non-JSON error body. - `OPTIONS` on a permitted collection returns a field-level metadata document (types, required flags, max lengths). With credentials this is the only way to read request schemas. - No idempotency key, no request-id header, no rate-limit headers, no `Retry-After`. ## Measurement surface - **Universal Site Pixel** — a first-party Google Tag Manager Community Gallery template (https://github.com/tvScientific/gtm_tag, Apache-2.0). Takes one parameter, `tvsci_tracking_id`, and injects `https://assets.tvspix.com/.js`. It must be the first script on the page, before any event-related calls. - **Conversion postbacks** run server-to-server into MMPs and affiliate networks (Adjust, AppsFlyer, Airbridge, Singular, Awin, Partnerize/Ascend, Avantlink, Rockerbox, Measured, Triple Whale, Invoca). The postback payloads are documented by those partners, not by tvScientific. ## Operations - Status page: https://status.tvscientific.com/ — machine-readable at https://status.tvscientific.com/api/v2/summary.json (plus status, components, incidents and scheduled-maintenances endpoints, and RSS/Atom history feeds). Seven components, dated incident history back to 2025. This is the most reliable programmatic signal tvScientific publishes. - Trust center: https://trust.tvscientific.com/ — client-rendered; no certification is verifiable anonymously. - Support is credentialed: help.tvscientific.com redirects to an Atlassian Service Desk login, support.tvscientific.com to a Salesforce Community sign-in. ## Human surfaces - Website: https://www.tvscientific.com - Platform: https://www.tvscientific.com/tvscientific-platform - Integrations: https://www.tvscientific.com/integrations - Partners: https://www.tvscientific.com/partners - Academy (learning): https://www.tvscientific.com/academy - Blog: https://www.tvscientific.com/insight - Press: https://www.tvscientific.com/press - Contact: https://www.tvscientific.com/contact - Demo request: https://www.tvscientific.com/ctv-demo - Application login: https://tvscientific.app - Platform Master Agreement: https://www.tvscientific.com/platform-master-agreement-demand/ - Privacy policy: https://www.tvscientific.com/privacy-policy/ - GitHub: https://github.com/tvScientific ## Provenance Generated 2026-08-12 by the API Evangelist enrichment pipeline from live anonymous probes and the tvScientific public web surface. Not published by tvScientific. Every claim above is recorded with its HTTP status in https://github.com/api-evangelist/tvscientific/blob/main/discovery/tvscientific-api-discovery.yml