Project stage: Live in production.
Where it stands now
Live in production with a publishing workflow connecting research, editorial review and reader-facing fact-check results.
- Reader-facing published fact-check results
- FactGuard: AI editorial review before publication
- Human research workflow with automated recovery and retries
“Jonathan doesn't take orders. He takes ownership. There's a difference and most founders never find someone who actually knows the difference. dBunk is a category-defining platform in a $50B+ media integrity market, and TDAT built every single piece of it. When I needed someone to go to war with, Jonathan showed up in full battle rattle.”
The starting point
dBunk needed a platform that connects everyday readers to a professional research team, fast enough for a consumer product, while keeping its editorial standards.
What TDAT built
The full platform: a real-time news feed, on-demand fact-check requests wired into the research team's workflow, subscription billing, an admin monitoring suite, and retry handling that reruns failed checks until they complete. After an editorial bug reached the live product, we shipped FactGuard, an AI editorial review layer that checks results before publication.
One feature: a fact-check request with a publishing path
The reader-facing request connects to a human research workflow and then to a published result. The platform also includes retry handling and FactGuard, an AI editorial review layer introduced to check results before publication.
The lesson for your build
Design the review path alongside the customer interface. For a similar system, define the evidence reviewers need, the conditions that block publication and the owner of failed work. An automated editorial check is one review step and does not by itself make every published claim correct.
Stuck somewhere similar?
Tell us what you built and where it is stuck. We will say what we would do next and what it would cost to find out.
Book a free consultation