Enveil

• Enveil's independently recorded buyers are federal, so its financial-services and healthcare pitch could stall if commercial budgets for data-in-use security are slow to form.
• The Common Criteria evaluation excluded the homomorphic encryption, so a buyer who requires a formal third-party evaluation of the cryptography itself finds none in the reviewed sources.
• The last priced venture round closed in April 2022 with no revenue or margin in the reviewed sources since, so a long commercial ramp could force Enveil to raise again on weaker terms.
• A large data platform could ship good-enough privacy-preserving computation and absorb the commercial use cases Enveil sells against.
• The commercial customers named in the reviewed sources are Enveil's own investors, so a bank evaluating it finds no unaffiliated reference in that record.

Enveil sells against a gap that ordinary encryption leaves open. Encryption protects data at rest and in transit, but searching or analyzing it has historically meant decrypting it first, which creates the moment of exposure Enveil calls the Data in Use problem. The company aims the capability at organizations that must use sensitive data without moving or exposing it, and it names financial services, healthcare, and the public sector as its verticals.

The framing is specific rather than generic. Enveil ties the problem to situations where data cannot be pooled, such as collaboration across jurisdictions and organizational silos, and its FAQ separates a usable commercial product from the homomorphic-encryption research it builds on. SecurityWeek reports that the core technology originated inside the National Security Agency, which corroborates where the capability came from rather than how large the problem is.

What the reviewed sources do not carry is a measure of the pain. No fetched source puts a figure on what organizations lose to data they cannot pool, and no regulation in the reviewed record requires this class of product. A buyer sizing this problem therefore has Enveil's framing and no independent figure. [s1, s4, s12, s13]

Enveil ships two products rather than a cryptography library. ZeroReveal Search runs encrypted search and watchlisting so a user can query data without revealing the query, and ZeroReveal Machine Learning runs encrypted evaluation, inference, and federated training so models and the data they read stay protected while they are used. Both arrive as a two-party client and server that sit as a proxy layer beside a customer's existing systems.

The company draws its own line between research and product. Its FAQ argues that homomorphic-encryption libraries supply basic cryptographic components while turning them into enterprise software takes engineering, algorithmic, and integration work, and it states plainly that Enveil does not sell proprietary cryptography. That engineering is the value Enveil claims to add.

Independent validation of the encryption is thinner than the certification suggests. The Common Criteria validation report for the ZeroReveal Compute Fabric records that the product was evaluated as a software application only, that the homomorphic encryption and the encryption and decryption of searches and results were outside the evaluation's scope, and that the evaluation makes no security claims about those features. A buyer weighing the cryptography has Enveil's own performance figures and no formal third-party test of them in the reviewed sources.

The nearest external check is a regulator's competition. At the Financial Conduct Authority's 2019 TechSprint, a judging panel assessing teams on criteria including effectiveness and potential market readiness gave second prize to a solution from an Enveil team that used homomorphic encryption for queries among banks. That is an outside body judging an implementation of the capability rather than an evaluation of the encryption itself. [s2, s3, s4, s18, s19]

Enveil competes in a small field of vendors commercializing homomorphic encryption, and the claim that the field is small belongs to Enveil. TechCrunch reported it as the company's own pitch, and a parenthetical in the same article names Duality, IBM, and Zama as others working to commercialize the technique. CB Insights lists a wider competitive set that also takes in Opaque, Anjuna, Omnisient, Cosmian, and Trust3 AI.

The independent placement Enveil holds is a middle one. CB Insights names it a Challenger in a confidential-computing landscape of sixteen companies that also includes IBM, Intel, and Anjuna, and SecurityWeek describes it as one of the more mature players in the privacy-enhancing-technologies category. Neither source puts Enveil at the front of the field.

A large data platform is the bigger long-term threat. One that shipped good-enough privacy-preserving computation could absorb the commercial use cases and leave Enveil's federal accreditation work as the part hardest to displace. No fetched source documents such a move. [s14, s15, s20]

Enveil's traction is documented most firmly in federal work. The federal award database lists an Army Phase II contract, number W51701-24-C-0171, worth $1,866,377, for encrypted model inference under Project Linchpin, and Enveil announced that same selection in 2024 as a direct-to-Phase-II award won through an xTech competition. Those are one engagement with two records. Enveil separately announced in 2025 a Defense Innovation Unit contract putting its software into the Hybrid Space Architecture effort with U.S. Space Force.

Commercial traction is asserted more than shown. At the 2022 raise the founder told TechCrunch that strategic investors including USAA, Mastercard, and Capital One Ventures were also paying customers, and the same article carries a 300% revenue increase since the 2020 Series A as the company's figure. Enveil's own Series B announcement, carried by Help Net Security, describes an existing customer base of global organizations in financial services, healthcare, and government.

What the reviewed sources do not carry is a named commercial customer unaffiliated with Enveil's investors. Enveil's financial-services page publishes a case-study summary that describes benefits without identifying a bank and invites visitors to request a detailed case study. The one independent record in the reviewed sources that names banks is the UK Financial Conduct Authority's 2019 TechSprint results page, which lists Enveil on the second-prize team with EY, BAE Systems, Refinitiv, HSBC, Barclays, and ING, a competition entry rather than a deployment. [s6, s7, s10, s15, s16, s17, s18]

The founder's background is the company's strongest documented asset. SecurityWeek reports that Ellison Anne Williams worked on the core technology inside the National Security Agency as a senior researcher for twelve years, and that when she left in 2015 she took the technology with her, exclusively. Enveil's About page adds that she has raised $40 million for the company.

The bench below her is named but less independently evidenced. Enveil lists CTO Ryan Carr, whose research the About page says has appeared in venues including Proceedings of the Royal Society, AAAI, and AAMAS, alongside functional executives across customer success, marketing, and sales. Those credentials appear on the company's own pages and in no fetched independent source, so a buyer weighing the depth of the team past the founder has only Enveil's account.

The founding story is consistent across sources. Enveil describes itself as founded in 2016 by U.S. Intelligence Community alumni with backgrounds in mathematics, algorithmics, and machine learning, which matches the press account of the technology transfer. [s1, s13]

Enveil's assurance collateral is built for government procurement. Enveil's product page presents the company as NIAP Common Criteria certified, a certification its announcement says its ZeroReveal solution achieved in 2018. Enveil describes the certification as validating a product against the security requirements for U.S. National Security System procurement across 28 member countries.

The certification covers less than the claim implies. The validation report records that the ZeroReveal Compute Fabric was evaluated as a software application only, that the homomorphic encryption and the encryption and decryption of searches and results were not in scope, and that the evaluation makes no security claims about those features. A buyer reading the certification as proof of the cryptography would be taking more from it than the report supports.

No SOC 2 report, ISO 27001 certificate, or trust portal appears on the probed surfaces. An enterprise buyer running a standard vendor review therefore has less to check than Enveil's federal accreditation suggests. [s5, s11, s19]

Add analyzed competitors to compare them side by side with Enveil.

Enveil works two buyer sets around one problem, the need to use sensitive data without moving or exposing it. One is national security, where defense users search and analyze data held across differing classifications, silos, and trust domains. The other is regulated commerce, where the company names financial services and healthcare and points at fraud prevention as the commercial use case.

Fraud is the named commercial use case. Enveil's ZeroReveal Search 6.0 announcement lists use cases running from fraud prevention to secure mission planning, and its financial-services page describes banks cross-matching and analyzing regulated data across silos. This is a narrow, high-assurance buyer set, and the engagements on the record are contracts.

Enveil ships two commercial products rather than a cryptography library. ZeroReveal Search runs encrypted search and watchlisting so a user can query data without revealing the query, and ZeroReveal Machine Learning runs encrypted evaluation, inference, and federated training so models and the data they read stay protected while they are used. Both run on homomorphic encryption and secure multiparty computation, delivered as a two-party client and server that act as a proxy layer and integrate without an architecture change.

The advantage Enveil claims is making that cryptography fast enough to use. The ZeroReveal Search 6.0 release added scalability work, cloud-hardware tuning, a smaller footprint for edge and Internet of Things devices, and scheduled queries. The throughput figure Enveil publishes is its own, and no fetched source tests it.

The pitch extends to protecting AI, and the market-position claim behind it belongs to Enveil. Gartner's 2025 market guide for AI trust, risk and security management identifies the company as a privacy-enhancing-technology provider using homomorphic encryption. TechCrunch reported as Enveil's own pitch the claim that it is one of a very small handful of companies commercializing homomorphic encryption, and a parenthetical in that same article names Duality, IBM, and Zama as others working to commercialize the technique.

Enveil's clearest go-to-market proof is federal, and it rests on one Army engagement recorded twice. The federal award database lists an Army Phase II contract, number W51701-24-C-0171, worth $1,866,377, for encrypted model inference under Project Linchpin, and Enveil announced that same selection in 2024 as a direct-to-Phase-II award won through an xTech competition. Enveil separately announced in 2025 a Defense Innovation Unit contract putting its software into the Hybrid Space Architecture effort with U.S. Space Force.

Commercial proof is the company's own account. The founder told TechCrunch that strategic investors including USAA, Mastercard, and Capital One Ventures were also paying customers, and the same article carries a 300% revenue increase since the 2020 Series A as Enveil's figure. Enveil's Series B announcement, carried by Help Net Security, describes an existing customer base of global organizations in financial services, healthcare, and government.

What the reviewed sources do not carry is a named commercial customer unaffiliated with Enveil's investors. The financial-services page publishes a case-study summary that identifies no bank and invites visitors to request a detailed case study. The one independent record in the reviewed sources that names banks is the UK Financial Conduct Authority's 2019 TechSprint results page, which lists Enveil on the second-prize team with EY, BAE Systems, Refinitiv, HSBC, Barclays, and ING, a competition entry rather than a deployment.

Enveil's fetched product and company pages disclose no prices and route interest to a demo request. The cited sources give no signal of the unit Enveil charges by. The one contract value on the record is the $1,866,377 Army Phase II award, a federal research contract rather than a commercial price.

Negotiated contracting fits the motion the record does show. Commercial-off-the-shelf software sold through federal contracting and multi-party enterprise deployments is more likely contracted than metered per seat, though the cited record does not establish whether Enveil prices per deployment, per use, per seat, or on some other basis.

Enveil delivers commercial-off-the-shelf software the customer runs in its own environment. The ZeroReveal client sits inside the enterprise and the server sits wherever the data lives, working alongside existing systems through standard interfaces, and Enveil states that the capability needs no change to system architecture, storage format, or application code.

Recent releases push toward constrained and federal settings. ZeroReveal Search 6.0 added a smaller footprint for edge and Internet of Things devices, scheduled queries, and scalability work, and the Defense Innovation Unit engagement integrates the software into government mission workflows. The work of adoption falls on wiring the software into a customer's data pipeline.

Enveil's assurance collateral is built for government procurement. Enveil's product page presents the company as NIAP Common Criteria certified, a certification its announcement says its ZeroReveal solution achieved in 2018. Enveil describes the certification as validating a product against the security requirements for U.S. National Security System procurement across 28 member countries.

The certification covers less than the claim implies. The validation report records that the ZeroReveal Compute Fabric was evaluated as a software application only, that the homomorphic encryption and the encryption and decryption of searches and results were not in scope, and that the evaluation makes no security claims about those features. A buyer reading the certification as proof of the cryptography would be taking more from it than the report supports.

No SOC 2 report, ISO 27001 certificate, or trust portal appears on the probed surfaces. An enterprise buyer running a standard vendor review therefore has less to check than Enveil's federal accreditation suggests.

Enveil sells a commercial product layer built on open privacy-enhancing-technology research rather than a library. Its FAQ draws the line explicitly: homomorphic-encryption libraries supply basic cryptographic components, and turning those components into enterprise-grade solutions takes engineering, algorithmic, and integration work. Enveil also states that it does not sell proprietary cryptography.

The product is built to fit into a customer's stack rather than to be a platform others build on. ZeroReveal installs as a proxy capability beside existing systems and reads data where it already sits, which keeps Enveil a component in someone else's architecture. That posture leaves an absorption risk: a large data platform that shipped good-enough privacy-preserving computation could bundle away the commercial case, and no fetched source documents such a move.

The founder's background is Enveil's strongest documented asset. SecurityWeek reports that Ellison Anne Williams worked on the core technology inside the National Security Agency as a senior researcher for twelve years, and that when she left in 2015 she took the technology with her, exclusively. Enveil's About page adds that she has raised $40 million for the company.

The bench below her is named but less independently evidenced. Enveil lists CTO Ryan Carr, whose research the About page says has appeared in venues including Proceedings of the Royal Society, AAAI, and AAMAS, alongside functional executives across customer success, marketing, and sales. Those credentials appear on the company's own pages and in no fetched independent source, so a buyer weighing the depth of the team past the founder has only Enveil's account.

The founding story is consistent across sources. Enveil describes itself as founded in 2016 by U.S. Intelligence Community alumni with backgrounds in mathematics, algorithmics, and machine learning, which matches the press account of the technology transfer.