Built by Bernard Huang for teams shipping AI-assisted content workflows.
I built VeracityAPI because AI flooded publishing, search, agent, and media pipelines with outputs that can look shippable while still being generic, unsupported, weakly sourced, or risky to trust. VeracityAPI is the linter for that moment: one API call that tells your system whether to allow, revise, queue for human review, or reject.
Bernard Huang

Bernard co-founded and bootstrapped Clearscope to 7-figure ARR. Clearscope has been used by content and growth teams at organizations including Nvidia, HubSpot, Adobe, IBM, Condé Nast, and more.
That background shaped VeracityAPI: practical content quality systems, measurable editorial workflows, and developer-first tools that help teams ship better content at scale.
LinkedIn · bernardjhuang.com · @bernardjhuang · hello@veracityapi.com
VeracityAPI · Austin, TX
VeracityAPI is content trust infrastructure for developers building agent-driven publishing, ingestion, moderation, and media review workflows.
- Text, image, audio, and video analysis are live.
- Customer-submitted content is not used to train models.
- Raw customer content and media are not stored by default.
- Outputs are workflow-routing signals, not forensic accusations.
Why now
AI made it cheap to generate plausible drafts, captions, sources, images, audio clips, and videos. The bottleneck moved from creation to trust: should this output be published, cited, indexed, trained on, or sent to a human? VeracityAPI exists for that routing decision.
Why this is not another authorship detector
Binary “human vs AI” labels are brittle and often the wrong product question. VeracityAPI focuses on observable workflow risk: generic phrasing, low specificity, weak provenance, unsupported claims, synthetic media cues, and missing review context.
Current scope
VeracityAPI is a production-focused API with a unified POST /v1/analyze endpoint, public docs, OpenAPI, SDKs, MCP support, llms.txt, agents.json, usage-based pricing, and transparent limitations.