Overview: super.AI is an Intelligent Document Processing (IDP) platform that converts raw documents into structured data using no-code workflows. The platform orchestrates AI models, software components, and a curated human-in-the-loop workforce (the Data Processing Crowd) to balance quality, cost, and speed. Key capabilities include document classification, field extraction and enrichment, validation against ERPs/databases, an interactive review UI, and export/push to downstream systems. Architecture highlights include task orchestration, an optimizer (to set targets for quality/cost/speed), a router (assigns work to AI/software/people), a combiner (merges results), and a trainer (continuous learning). The platform supports OCR, handwriting (with caveats noted), scanned PDFs and common image formats. Claims & metrics: The site cites accuracy and outcomes in the range of ~91–99%+ and customer cases reporting up to 96–100% automation for some customers, along with large ROI/cost-savings claims (examples in marketing material reference figures like $9M annual savings and >$100M claimed savings across customers). Typical benefit claims include ~75% reduction in processing time and ~80% reduction in data-entry costs; these are marketing/testimonial statements and should be validated in a POC or trial. Pricing & trials: Pricing is consultative and tailored (usage/volume-based). There are no public fixed plans or price tables on the site. Pricing factors include annual document volume, number of fields to extract, customization/training needs, SLAs, and integrations. A trial is available via the Store page: processing up to 100 pages over 7 days, but trials are set up by scheduling a meeting (trial accounts are not self-provisioned directly on the site). To get a quote or demo, use the Contact / Request Demo flow to schedule a personalized demo and pricing review. Products & store: Store items/apps referenced include "Extract" (field extraction templates / General Document Processor) and "Classify" (document classifier app). The extraction approach is template-agnostic / LLM-based. Supported languages listed on-site include English, French, German, Norwegian, Spanish, Dutch, Polish, Portuguese and Italian; the company invites contact for less-common languages or non-Latin scripts. Docs & API: Documentation includes Quick Start, New User Guides, Advanced Guides, API Reference and FAQs, with API endpoints/parameters and privacy/security guidance. A public API domain and API docs are referenced on the site. Integrations & deployment: The site references pre-built plug-and-play integrations with ERP and Document Management Systems, and custom integrations available via consulting. Deployment options described include public or private cloud deployments. Human-in-the-loop & quality: The Data Processing Crowd is a curated, on-demand workforce for labeling, post-processing, exception handling and QA, used to reach guaranteed quality thresholds. The platform references extensive QA mechanisms (150+ QA checks mentioned) and configurable thresholds that trigger human review for low-confidence results. Support & trust: Customer support is described as tiered, with access to a support library, on-demand technical support, dedicated technical account managers for higher tiers, varying SLAs, Slack/video chat, 24x7 docs, optional live training, and integration engineering support. Trust and privacy content highlights SOC 2 and GDPR alignment, sub-processor transparency, encryption, access controls, and retention policies. The privacy policy identifies Super.AI Inc. (Delaware) and a German affiliate (Canotic GmbH) as controllers and provides contact points; security and legal contact emails are listed in policy content. Limitations & notes: No public price table or self-serve subscription tiers were found — pricing is custom and consultative. The trial exists but appears to require contacting sales and scheduling (up to 100 pages for 7 days as indicated on the Store page). Site content is marketing-oriented; accuracy and ROI claims are customer/testimonial/marketing statements and should be validated in a POC or trial. Suggested next steps: 1) Book a demo via the contact page to request the 7-day 100-page trial and obtain a tailored quote based on annual volumes and extraction fields. 2) Run a trial/POC with sample documents to validate accuracy and automation rates on your document types. 3) Request SOC 2 report, DPA, and sub-processor/security details during sales/contract discussions if required. 4) Review the API docs and Quick Start to estimate integration effort and throughput. 5) Ask for a sample SLA/pricing scenario mapping desired quality/cost/speed targets to expected human/AI mix and costs. Notes on sourcing: All information in this description is taken from a site review summary and reflects content/claims presented on the super.AI site and associated documentation; claims and marketing outcomes should be validated through trial/POC and contractual review.