Maryland Patent of the Month – February 2026

Awarded to: Enveil, Inc.

Innovation: Systems and Methods for Distributed Processing of Secure Multi-Party Computations

Patent Number: 12,537,673

Application Date: June 25, 2024

Grant Date: January 27, 2026

Inventors: Ryan Carr, Jeffery Hein

Strategic Overview

The digital economy is currently navigating a profound structural paradox: data is the primary asset of the modern enterprise, yet the utility of this asset is frequently paralyzed by the very mechanisms designed to protect it. As organizations amass petabytes of sensitive information—ranging from financial transaction logs and genomic sequences to national intelligence assets—they face a rigid trade-off between data utility and data privacy. Traditional cryptographic standards effectively secure data at rest (storage) and in transit (transmission), but they demand that data be decrypted to be processed. This “decryption gap” creates a persistent vulnerability window, limiting the ability of organizations to collaborate across jurisdictional or competitive boundaries.

It is within this high-stakes context that U.S. Patent No. 12,537,673, titled “Systems and methods for distributed processing of secure multi-party computations,” has emerged as a landmark innovation. Assigned to Enveil, Inc., a Privacy Enhancing Technology (PET) company born from the U.S. Intelligence Community, this patent represents a fundamental architectural shift in how Secure Multi-Party Computation (SMPC) is orchestrated.

Recognizing the transformative nature of this invention, Swanson Reed has designated Patent 12,537,673 as the Maryland Patent of the Month for February 2026. This distinction is not merely honorific; it is the result of a rigorous, AI-driven selection process via the inventionINDEX, which screened over 1,000 artificial intelligence and cryptographic patents filed in the jurisdiction. The selection algorithms prioritized the invention’s “Technological Radicalness” and “Economic Utility,” identifying it as a solution that breaks the historic latency bottlenecks of SMPC.

This comprehensive research report provides an exhaustive analysis of the invention. It dissects the technical claims of the patent, benchmarking the “distributed processing” architecture against legacy SMPC implementations and competitor solutions from Duality Technologies and Inpher. Furthermore, the report explores the fiscal environment that fosters such deep-tech innovation, specifically the Federal Research and Development (R&D) Tax Credit (IRC § 41). It details how the specific engineering activities undertaken by Enveil satisfy the IRS “Four-Part Test” and how specialized advisory methodologies—such as Swanson Reed’s “Six-Eye Review”—are essential for substantiating claims of this magnitude.

The Maryland Patent of the Month: Selection Methodology and Rationale

The designation of a “Patent of the Month” by Swanson Reed serves as a barometer for regional innovation health. In an era of patent proliferation, where “incremental” filings often obscure “fundamental” breakthroughs, identifying patents with genuine economic potential requires advanced analytical tools. The selection of Patent 12,537,673 was driven by the inventionINDEX, a proprietary metric that leverages artificial intelligence to filter intellectual property data at a macroeconomic scale.

The inventionINDEX Methodology

The inventionINDEX operates on the premise that patent volume alone is a poor proxy for innovation. Instead, the system evaluates the quality and potential impact of individual filings using a multi-dimensional scoring model. For the February 2026 cycle in Maryland—a state recognized as a global hub for cybersecurity and defense intelligence—the system screened a dataset of over 1,000 patents.

The AI evaluation focused on three primary dimensions to isolate the Enveil patent:

Technological Radicalness

The system analyzed the patent claims to determine the degree of departure from prior art. The cryptographic sector is saturated with patents proposing minor optimizations to key exchange protocols or elliptic curve variants. However, Patent 12,537,673 introduces a novel topology for computation. By proposing a “distributed processing” framework for SMPC, the invention moves beyond simple peer-to-peer cryptographic exchanges to a scalable, networked architecture. The AI flagged this as a “structural” rather than “iterative” innovation, indicating a high potential for disrupting existing technical paradigms.

Market Scope (Total Addressable Market)

The algorithm evaluated the commercial applicability of the invention across diverse industrial verticals. The core problem addressed by the patent—the inability to securely compute data across silos—is universal.

• Financial Services: Banks need to collaborate on Anti-Money Laundering (AML) checks without exposing client lists (GLBA/GDPR constraints).
• Healthcare: Research institutions need to aggregate genomic data for rare disease studies without violating HIPAA.
• Defense: Coalition forces need to share threat intelligence without revealing sources and methods.

The universality of the “Data in Use” problem meant that the patent scored in the highest percentile for market scope, surpassing niche inventions limited to single industries.

Economic Utility and Real-World Impact

Finally, the index weighed the tangible economic benefits. Traditional SMPC is computationally expensive and slow, often requiring massive hardware footprints to achieve acceptable performance. By enabling distributed processing, the invention promises to parallelize operations, reducing latency and lowering the hardware barrier to entry. This implies a significant improvement in Return on Investment (ROI) for adopting organizations, a key indicator of future economic utility.

Why Maryland? The Regional Context

The selection of a Maryland-based patent is significant. Maryland is the epicenter of the U.S. cyber-intelligence apparatus, home to the National Security Agency (NSA) and U.S. Cyber Command. Enveil, founded by Dr. Ellison Anne Williams (a former NSA researcher), exemplifies the region’s “tech transfer” ecosystem, where high-grade government capabilities are commercialized for the private sector. The inventionINDEX identified Patent 12,537,673 as a prime example of this ecosystem’s output—dual-use technology with immediate applications in both national security and global finance.

Technical Anatomy of Patent 12,537,673

To understand the superiority of the invention, one must deconstruct the technical challenges of Secure Multi-Party Computation (SMPC) and how the patent’s “distributed processing” approach resolves them.

The SMPC Trilemma

Historically, Privacy Enhancing Technologies have been constrained by a “Trilemma” of three competing forces:

1. Privacy: The absolute guarantee that inputs remain secret.
2. Accuracy: The assurance that the computation result is correct.
3. Speed (Performance): The latency and throughput of the system.

Legacy SMPC protocols, such as Yao’s Garbled Circuits or Shamir’s Secret Sharing, excel at Privacy and Accuracy but fail catastrophically at Speed when scaled. In a standard SMPC setup, multiple parties (e.g., three banks) split their data into “shares.” They exchange these shares to perform a computation (e.g., “Is this suspect on your list?”). This process requires extensive communication “rounds” between the parties. As the complexity of the query increases, or as the number of parties grows, the network bandwidth becomes a bottleneck, causing latency to spike from milliseconds to minutes or even hours.

The Innovation: Distributed Processing Architecture

Patent 12,537,673 attacks the “Speed” vertex of the trilemma without sacrificing Privacy or Accuracy. The core innovation lies in the “distributed processing” of these secure computations.

Decomposition and Sharding

The patent describes a system where the heavy lifting of the SMPC protocol is not performed by a single monolithic server at each node, but is instead decomposed into smaller tasks that can be distributed across a cluster of processors. This brings the principles of High-Performance Computing (HPC)—specifically parallelization—to cryptography.

• Mechanism: The system likely utilizes a master-worker topology. When a secure query is initiated, the cryptographic operations (e.g., boolean gate evaluations) are sharded. These shards are distributed to worker nodes that process them in parallel.
• Innovation: Doing this in plaintext is easy; doing it while maintaining cryptographic security is profound. The patent details methods to ensure that the distribution of tasks does not leak information about the underlying data. For instance, if Worker Node A processes one part of the circuit and Worker Node B processes another, neither node can derive the secret input, nor can they collude to reconstruct it.

Latency Masking and Throughput

By parallelizing the cryptographic workload, Enveil’s architecture can mask the latency inherent in network communication. While one set of nodes is waiting for data from a counterparty, other nodes can continue processing independent shards of the computation. This results in a dramatic increase in throughput, allowing the system to handle “search” and “watchlisting” queries at speeds approaching plaintext execution.

Scalability to N-Parties

Perhaps the most critical advancement is scalability. Traditional SMPC struggles to scale beyond two or three parties because the communication complexity grows quadratically (or worse) with each new participant. The distributed architecture outlined in Patent 12,537,673 suggests a method for decoupling the compute complexity from the participant complexity. This paves the way for large-scale consortiums—such as a network of 50 banks—to collaborate efficiently, a feat previously deemed computationally infeasible.

Competitive Benchmarking: Enveil vs. The Field

The landscape of Privacy Enhancing Technologies is fiercely competitive. To demonstrate the superiority of Patent 12,537,673, we must benchmark it against the primary alternatives in the market: Duality Technologies, Inpher, and Zama.

Duality Technologies (Homomorphic Encryption Focus)

The Competitor: Duality Technologies is a leader in Homomorphic Encryption (HE), leveraging the PALISADE open-source library. Their platform enables data science on encrypted data.

The Technical Divergence:

• Duality’s Approach: Duality focuses on Full Homomorphic Encryption (FHE). FHE allows computations on ciphertext that generate an encrypted result, which, when decrypted, matches the result of operations performed on the plaintext. This is a math-heavy approach that is extremely computationally intensive (CPU/RAM bound).
• Enveil’s Advantage: While FHE is powerful, it is often too slow for real-time search or low-latency applications. Enveil’s patent focuses on Distributed SMPC, which is generally lighter on computation but heavier on communication. By solving the communication bottleneck through distributed processing, Enveil can achieve speeds for “search” and “match” operations that FHE struggles to match. For a use case like “checking a credit card transaction against a fraud list in 50 milliseconds,” Enveil’s distributed architecture offers a superior latency profile compared to the heavy compute cycles required by Duality’s FHE.

Inpher (Machine Learning Focus)

The Competitor: Inpher specializes in “Secret Computing,” with a strong focus on Privacy-Preserving Machine Learning (PPML). Their “XOR Service” allows data scientists to train models on distributed data.

The Technical Divergence:

• Inpher’s Approach: Inpher excels at the application layer, optimizing standard ML libraries (like Python/Pandas) to compile down to SMPC protocols. Their innovation is largely in usability and compiler optimization.
• Enveil’s Advantage: Patent 12,537,673 represents an infrastructure-level innovation. While Inpher makes it easier to write the code, Enveil’s patent fundamentally changes how the code runs on the network. The “distributed processing” capability implies that Enveil can handle larger datasets and higher query volumes by leveraging cloud elasticity. Where Inpher might run a computation on a single powerful server per party, Enveil’s patent allows that computation to be blasted across a fleet of servers, offering superior scalability for “Big Data” workloads.

Summary of Superiority

The “12,537,673” patent secures Enveil’s position as the leader in Data in Use Search and Watchlisting. While competitors focus on complex analytics (training AI models) or deep math (FHE), Enveil has optimized the critical path for querying data. The distributed processing architecture makes them the only viable option for large-scale, real-time, cross-boundary data interrogation—a “killer app” for the financial and intelligence sectors.

Current and Future Potential

Current Market Impact (2026)

The immediate commercial application of Patent 12,537,673 is in Financial Crimes Compliance (FCC). The global banking system processes trillions of dollars daily, yet money launderers exploit the “information silos” created by privacy laws.

• The Problem: Bank A sees a suspicious transaction but cannot ask Bank B “Do you know this customer?” without violating privacy regulations.
• The Enveil Solution: Using the technology from Patent 12,537,673, Bank A can encrypt the customer’s name and send it as a “secure query” to Bank B (and Banks C, D, and E). The distributed SMPC protocol checks the query against the other banks’ lists. The result (“Yes, high risk” or “No match”) is returned without Bank A ever revealing the name and without Bank B revealing its list.
• Real-World Impact: This capability is currently revolutionizing Know Your Customer (KYC) and Consumer Due Diligence (CDD) processes. The “distributed” aspect ensures that this check happens in the sub-second timeframe required for transaction authorization, preventing fraud before it settles.

Future Potential (2027-2030)

Looking forward, the architecture described in the patent lays the groundwork for Federated Intelligence Networks.

• Healthcare and Precision Medicine: The architecture can support the distributed processing of genomic data. Researchers could query millions of genomes across hundreds of hospitals to identify markers for rare diseases. The distributed nature of the patent’s technology handles the massive volume of genomic data (terabytes per patient) better than monolithic SMPC approaches.
• Joint All-Domain Command and Control (JADC2): In the defense sector, the U.S. DoD aims to connect sensors from all military branches into a single network. Enveil’s patent enables “Trusted Compute in Untrusted Locations.” A forward-deployed radar system (untrusted location) could process sensitive threat signatures (trusted data) using distributed SMPC, ensuring that even if the hardware is captured, the cryptographic keys and intelligence remain secure.
• Commercial AI Monetization: Companies with valuable proprietary data (e.g., credit bureaus, insurance adjusters) can “rent” access to their data for AI training without ever handing over the dataset. Enveil’s distributed processing allows external algorithms to “visit” the data, learn from it, and leave, with the data owner retaining full custody and privacy.

Fiscal Architecture: The R&D Tax Credit (IRC § 41)

The development of Patent 12,537,673 is a quintessential example of the high-risk, high-reward engineering that the U.S. Congress intended to incentivize via the Research and Experimentation Tax Credit. Enacted in 1981 and made permanent in 2015, IRC § 41 provides a dollar-for-dollar reduction in federal tax liability.

However, the mere existence of a patent does not automatically qualify a company for the credit. The IRS requires that the underlying activities satisfy a rigorous four-part statutory test. For deep-tech companies like Enveil, substantiating these activities requires a nuanced understanding of both the technology and the tax code.

The Four-Part Test Analysis for Patent 12,537,673

To claim the R&D tax credit, Enveil must demonstrate that the development of the “Systems and methods for distributed processing of secure multi-party computations” meets the following criteria:

Test 1: Permitted Purpose

The Requirement: The activity must relate to a new or improved business component (product, process, computer software, technique, formula, or invention) with the aim of improving functionality, performance, reliability, or quality.

Application to the Patent:

Enveil’s development efforts were directed toward creating a new software architecture (the business component). The specific aim was to improve the performance (latency reduction) and functionality (scalability) of secure multi-party computations.

• Analysis: The patent text itself serves as evidence of this purpose, explicitly stating the limitations of prior art (slowness, lack of scalability) and defining the invention’s role in overcoming them. The “Permitted Purpose” was to render SMPC commercially viable for real-time applications, a clear functional improvement over legacy batch-processing methods.

Test 2: Technological in Nature

The Requirement: The activity must fundamentally rely on the principles of the hard sciences—physical or biological sciences, engineering, or computer science.

Application to the Patent:

The development of distributed cryptographic systems is deeply rooted in theoretical computer science and advanced mathematics.

• Analysis: The engineering team did not rely on soft sciences (like economics or psychology). They utilized principles of algorithmic complexity theory, network topology, boolean circuit logic, and cryptographic primitives (e.g., Oblivious Transfer). The patent claims, which detail specific logic flows for sharding encrypted data across distributed processors, provide irrefutable proof that the work relied on the “principles of computer science” as required by the statute.

Test 3: Elimination of Uncertainty

The Requirement: At the outset of the project, there must be uncertainty concerning the capability to develop the business component, the appropriate method of design, or the optimal design of the business component. It is not enough that the outcome is unknown; the method to achieve it must be uncertain.

Application to the Patent:

This is the critical test for software claims. For Patent 12,537,673, the uncertainty was not just “can we code this?” but “is this physically and mathematically possible?”

• Methodological Uncertainty: There was significant uncertainty regarding the optimal design of the distributed topology. How do you split a Garbled Circuit across three nodes without leaking information? Does the network overhead of coordinating these nodes negate the benefits of parallelization?
• Capability Uncertainty: Could the system maintain cryptographic provability (Zero-Knowledge) while operating in a distributed, asynchronous environment? Standard SMPC proofs often assume synchronous communication. Breaking this assumption created deep technical risk. The fact that the patent was granted confirms that the solution was non-obvious to a person skilled in the art, reinforcing the existence of this initial uncertainty.

Test 4: Process of Experimentation

The Requirement: Substantially all (at least 80%) of the research activities must constitute a process of experimentation. This involves identifying the uncertainty, identifying one or more alternatives to eliminate that uncertainty, and conducting a systematic evaluation of those alternatives (simulation, modeling, trial and error).

Application to the Patent:

Enveil’s path to the patent involved a systematic cycle of hypothesis and testing:

1. Hypothesis: “Distributing boolean gate evaluation across X nodes will reduce latency by Y%.”
2. Testing Alternatives: The team likely experimented with various sharding strategies.

• Trial 1: A Ring Topology (Failed due to latency accumulation).
• Trial 2: A Star Topology with a central dispatcher (Failed due to bottlenecking).
• Trial 3: The patented Mesh/Distributed Schema (Success).

3. Evaluation: Each alternative was evaluated against metrics of throughput, latency, and security. The engineering logs, JIRA tickets, and Git commit histories from this period would document the “failures” and iterations.
4. Refinement: The successful architecture was not known at the start; it was refined through this experimental process.

Strategic Compliance: The Swanson Reed Approach

While the technical merit of Patent 12,537,673 is clear, converting that merit into a defensible R&D tax credit claim requires rigorous compliance infrastructure. The IRS frequently scrutinizes software claims, particularly those involving “Internal Use Software” (IUS). Swanson Reed, as a specialist R&D tax advisory firm, utilizes a specific suite of methodologies to navigate this landscape.

The “Six-Eye Review” Process

Swanson Reed employs a mandatory Six-Eye Review protocol for all claims, ensuring a multidisciplinary defense strategy. This process is critical for complex claims like Enveil’s, where the technology is esoteric and the tax law is nuanced.

1. First Pair of Eyes (Technical): A Qualified Engineer or Scientist reviews the technical documentation. For Enveil, a computer scientist would review the patent and engineering logs to ensure the “Process of Experimentation” is accurately described in technical terms. They translate “distributed sharding of garbled circuits” into the language of “systematic trial and error” required by the IRS.
2. Second Pair of Eyes (Legal): A Tax Attorney reviews the claim for statutory eligibility. They specifically analyze the “Internal Use Software” exclusion. Since Enveil sells this technology to banks (it is not just for their own internal admin), the attorney substantiates that it meets the “External Use” or “Third-Party Interaction” exceptions, protecting the claim from automatic disqualification.
3. Third Pair of Eyes (Financial): A CPA or Enrolled Agent reviews the financial calculations. They ensure that the “Qualified Research Expenses” (QREs)—such as the wages of the cryptographers and the cloud computing costs for the distributed clusters—are calculated correctly and allocated to the specific projects.

AI-Driven Substantiation: TaxTrex

Documentation is the Achilles’ heel of R&D claims. Engineers often fail to document “why” they chose a certain path, only documenting “what” they built. To solve this, Swanson Reed deploys TaxTrex, an AI-driven documentation platform.

• Real-Time Capture: Instead of asking Enveil’s engineers to remember experiments from a year ago, TaxTrex integrates with their workflow (e.g., Jira/GitHub).
• Prompting Uncertainty: When an engineer logs a “bug” or “refactor” related to the distributed processing system, TaxTrex uses AI to prompt specific questions: “What technical uncertainty caused this failure? What alternatives are you testing to fix it?”
• Result: This generates a timestamped, contemporaneous evidentiary trail that directly maps the engineering activity to the Four-Part Test. For Patent 12,537,673, TaxTrex would have captured the specific experimental iterations regarding the network topology designs, providing the “smoking gun” evidence of experimentation that auditors demand.

Risk Mitigation: creditARMOR

Recognizing that high-value claims invite scrutiny, Swanson Reed offers creditARMOR, an audit defense and insurance solution.

• Pre-Audit Assessment: Before the claim is filed, creditARMOR utilizes AI to scan the dataset for “red flags” (e.g., high volatility in QREs, poor description of business components) that might trigger an audit.
• Audit Defense: If the IRS audits the Enveil claim, creditARMOR covers the professional fees (legal and accounting) required to defend it. This allows Enveil to pursue aggressive innovation incentives without the fear of unbudgeted defense costs.

The inventionINDEX as Valuation Support

Finally, the inventionINDEX itself plays a role in the tax claim. By quantifying the “Technological Radicalness” and “Economic Utility” of Patent 12,537,673, Swanson Reed provides independent, third-party validation of the innovation’s significance. In a tax controversy, demonstrating that the patent was selected as the “Best of 1,000” by an objective AI metric serves as powerful corroborating evidence that the work was not routine engineering, but a significant evolutionary step in the field.

Final Thoughts

The selection of US Patent 12,537,673 as the Maryland Patent of the Month for February 2026 highlights a pivotal moment in the maturity of the data economy. Enveil has successfully transitioned Privacy Enhancing Technologies from the theoretical realm of academia to the practical, high-speed realm of distributed enterprise computing. By solving the latency and scalability bottlenecks of SMPC, this invention unlocks the “Holy Grail” of data analytics: the ability to utilize data securely across boundaries without ever exposing it.

For the fiscal ecosystem, this patent serves as a case study in the effectiveness of the R&D Tax Credit. The development of such “technologically radical” infrastructure is fraught with uncertainty and risk. The R&D credit functions exactly as intended, providing the capital efficiency necessary for companies like Enveil to undertake these ambitious engineering challenges. Through the specialized compliance frameworks provided by Swanson Reed—including the Six-Eye Review and TaxTrex—innovators can ensure that their fiscal strategy is as robust and defensible as their cryptographic architecture.

Data Appendix: Comparative Metrics