Launched this week
Cogneris
Document AI API: extraction, validation, review, audit trail
2 followers
Document AI API: extraction, validation, review, audit trail
2 followers
Cogneris turns PDFs, scans and images into validated JSON you can defend in an audit: typed fields with per-field confidence and citations, validation rules, review queues, webhooks and a full audit trail. REST API, SDKs, MCP. 400 free credits, no card.
Free Options
Launch Team
RiverynLive like a hunter-gatherer. Keep the iPhone.
Promoted
Hi Product Hunt. We built Cogneris for the documents where a wrong field costs real money: payslips, bank statements, invoices, contracts, claims. Every extracted field comes back as typed JSON with a confidence score and a citation to where it came from, validation rules and a review queue catch the low-confidence ones before they reach your system of record, and an audit trail records the model version, prompt version, reviewer and validation status behind each value. It is API-first (OpenAPI 3.1, SDKs on npm, PyPI and NuGet, an MCP endpoint for your editor) with a customer-facing portal for intake. Pay-per-page, 400 free credits, no card. We publish our comparisons against Mindee, Extend, Nanonets and others on the site, including when they are the better fit. Happy to answer anything about accuracy, audit evidence or pricing.