Veriff

Market position

Veriff positions itself as an enterprise-scale identity verification platform that owns its core models rather than reselling third-party technology, contrasting itself with orchestration-layer vendors [0][1]. Its API has been compared to Stripe [7]. Investors and executives characterized it in 2021-2022 as an industry-leading IDV product, and the 2021 Series B was described by the company as the largest Series B raised in the IDV market to that date [5][6]. Estonian venture capitalists named Veriff startup of the year at the end of 2018 [7].

Veriff emphasizes that it develops its core decision engine, liveness and face match, and document forensics models in-house and trains them on its own data, in contrast to vendors that license and rebrand third-party models; adjacent checks such as AML and database verification are provided by named specialist partners disclosed in its sub-processor list [0][1]. Other stated differentiators are a video-first verification approach combined with data cross-linking, analysis of more than 1,000 technological and behavioural data points, passive liveness that requires no additional user actions, and broad document and language coverage [6][4][7].

Technology

Veriff combines AI-based automated decisioning with human-in-the-loop review [4][2]. The decision engine analyzes more than 1,000 data points including network and device analytics, plus liveness and biometric checks derived from selfies and optional background video [4]. The company states it supports more than 12,500 government-issued ID specimens across 230+ countries and territories, in Arabic, Hebrew, Chinese, Cyrillic, Japanese, Korean and Latin-based scripts, and localizes end-user flows into 50 languages and dialects [0][1][4]. Assisted Image Capture uses machine learning to give real-time capture feedback on issues such as poor lighting, glare or a cropped ID, and passive liveness detection avoids gesture-based prompts [0][1][4]. Reinforcement learning from human feedback is used to add newly issued document specimens to the database [4]. Veriff cross-links session data with other sessions to detect systematic fraud, and released a near-field communication tool in October 2019 to read data directly from biometric document chips [7]. The company states it builds core models — decision engine, liveness and face match, and document forensics — in-house, while using named specialist partners for adjacent checks such as database verification and AML screening [0][1].

Go-to-market

Direct enterprise sales supported by demo requests and "talk to us" contact flows, a free trial, self-serve developer resources (documentation, sandbox, SDKs) and a live product demo [0][1][3][4]. Following its Series C the company said it would grow sales and marketing teams to onboard more global clients and build on customer success programs [5]. It has also made its service available to startups and SMBs through a Zapier integration [5].

Online businesses that must confirm user identity, meet KYC/KYB/AML obligations, or block fraud; the site states more than 3,000 businesses are customers [0][1]. Named or described sectors include financial services, banking, cryptocurrency, gaming, healthcare, fintech and mobility [5][6][7]. Publicly identified users include Estonian retailer Hansapost, Estonian bank LHV, Berlin transport operator Berliner Verkehrsbetriebe (for its Jelbi mobility app), Blockchain.com, Bolt, Opera, Mintos, Turo, Grow Mobility, Verify 365 and Wise [7].

Geography

Founded and headquartered in Tallinn, Estonia, serving customers globally [7]. Veriff opened its first foreign office in New York City in November 2019, and in March 2020 said it was establishing customer-facing and account management teams in London [7]. As of 2021 the company reported offices in the U.S., UK and Estonia [6]. Document coverage spans 230+ countries and territories and business register coverage spans 300+ jurisdictions [0][1][4].

History

Veriff launched its service in Estonia in June 2016 and employed 12 people by September 2016. It worked for a year with Estonia's Ministry of Finance, Ministry of the Interior, the Estonian Data Protection Inspectorate and the e-Residency program on procedures for using national registries, with online retailer Hansapost among its first clients. Estonian bank Inbank (formerly Cofi) was its initial investor in 2016, and bank LHV began offering video-link account opening in 2017 [7]. Veriff graduated from Y Combinator in June 2018 and was named startup of the year by Estonian venture capitalists at the end of that year. In January 2019 it acquired Estonian network intelligence company BrowserID to extend fraud prevention beyond facial recognition and document checks. Agreements with Berliner Verkehrsbetriebe (June 2019) and Blockchain.com (August 2019) followed, and subscription services for SMBs launched in September 2019. The New York office opened in November 2019, by which point the company employed over 300 people in Tallinn and had raised EUR 7.35 million. Kotkas was named EY Estonian entrepreneur of the year in March 2020. In May 2020 Veriff laid off 63 employees as COVID-19 shifted volumes toward fully automated markets, and in June 2020 it raised a further EUR 13.8 million with headcount at 230. A $69M Series B led by IVP and Accel closed in April 2021, followed by a $100M Series C in January 2022 that took total funding to $200M [7][6][5].

Risks & controversies

In May 2020 Veriff announced the layoff of 63 employees, attributing the cut to COVID-19 shifting verification volume toward markets served fully automatically and reducing demand in markets requiring more human oversight [7]. Sources report differing valuations for the 2022 round ($1.5B per the company, $1.3B per Wikipedia) [5][7], and differing document/country coverage figures across time (9,000 IDs from 190 countries and 12.5K variations from 280 countries both cited in 2021 material; 12,500+ documents across 230+ countries on the current site) [6][0]. The website displays a notice that the service is not available in some visitor locations [0][1]. Verification requires camera and, in some client-configured flows, microphone access and session video/sound recording [2].