Rossum

Rossum is an AI-native intelligent-document-processing platform founded in 2017 in Prague by Tomáš Gogar, Petr Baudiš and Tomáš Tunys, and now headquartered in London with R&D still anchored in Prague. Its positioning against the broader IDP market is generational: where ABBYY is legacy OCR with templates, and Hyperscience is the "modern" rung built on CNN-and-template extraction with human-in-the-loop, Rossum was designed from inception as template-free, transformer-first capture — no zone definitions, no document-type training scaffolding, no OCR pre-processing layer.

What Rossum actually is. A cloud-native IDP platform targeting transactional document workflows — invoices, purchase orders, bills of lading, delivery notes, remittances, and for regulated verticals, claims forms, policy documents, and underwriting submissions. The core engine is a proprietary deep-learning model combining computer vision and natural-language processing that reads documents end-to-end rather than extracting fields from a predefined layout. This is what "template-free" means in practice: a new supplier, a new broker, a new claims form arrives and the model extracts without a human configuring zones. Rossum layers a human-in-the-loop validation UI on top, and exposes data through an API, a "Rossum Store" of pre-built connectors (SAP, Oracle, NetSuite, Workday, UiPath), and webhooks.

Funding and scale. Rossum raised a $100M Series A in October 2021 led by General Catalyst, with LocalGlobe, Seedcamp, Miton and Elad Gil participating — one of the largest Series A rounds in Eastern European tech history at the time. Cumulative disclosed funding is approximately $104M across three rounds. No Series B has closed as of April 2026. Publicly reported 2024 revenue was $44.9M with 167 employees; headcount is now around 188 across Europe, North America and Asia. Rossum remains private and independent; there is no acquisition on public record.

Analyst position. The strongest validation is a Leader placement in the IDC MarketScape: Worldwide Intelligent Document Processing Software 2023-2024 Vendor Assessment, citing "human-centric automation, orchestration and integration, and productivity for the large enterprise." In the 2025 Gartner Magic Quadrant for IDP (the first ever for this category, 18 vendors evaluated), Rossum was positioned as a Challenger — strong Ability to Execute but trailing Hyperscience, ABBYY and Automation Anywhere on Completeness of Vision. Everest Group has placed Rossum as a Major Contender on its IDP PEAK Matrix (2022, after Star Performer status in 2021); it did not make the Leader band in the 2025 Everest assessment.

Insurance footprint — handle with care. Rossum publishes an insurance solution page and markets the platform for claims intake, policy documents and underwriting submission extraction. However, the named-customer evidence is concentrated outside insurance: PepsiCo, Siemens, Bosch, Veolia, Cushman & Wakefield, COSCO Shipping, Delivery Hero, Deloitte, EY, Molson Coors, Suez, Eurowag, and, in financial services, Master Trust Bank of Japan. No P&C or life insurance carrier case study surfaced in April 2026 public search. A Phidea reader evaluating Rossum for an insurance submission-intake or claims-document use case should ask the vendor for direct carrier references rather than rely on the public case-study library.

Where it sits in old to new to AI. Legacy carriers and AP teams run ABBYY FlexiCapture, Kofax (now Tungsten) or offshore data entry. Hyperscience is the "modern" rung with CNN-plus-template and enterprise analyst depth. Rossum is the AI-native rung that arrived before the LLM era but was already architected around transformers — which explains why its product reads as more cohesive than Hyperscience-with-bolted-on-LLM, but also why newer LLM-wrapper entrants (Indico, Instabase, Roots, and general-purpose foundation-model APIs) can credibly compete on accuracy without Rossum's enterprise depth.