The modern identity check is designed to feel like nothing. A face fills an oval. A finger lands on glass. A camera blinks, and a bank account opens or a border gate clicks. The user sees one gesture. Aware, Inc. sees a sequence: capture the signal, grade its quality, look for a spoof, format the data, compare it with a stored template, apply policy and return a decision before impatience sets in.
That unglamorous machinery is the business. From Burlington, Massachusetts, a team of roughly 64 people builds software for recognizing faces, fingerprints, irises and voices. Its customers include government agencies, law-enforcement and border programs, financial institutions, healthcare organizations and enterprise security teams. Some buy directly. Many encounter Aware through a systems integrator or equipment maker that has tucked its code inside a larger identity system.
Aware is neither a startup nor a household name. It was founded in 1986 as what its history calls a “mathematical engineering company,” went public on Nasdaq in 1996 and once made money from DSL technology. By 2019 it had narrowed itself entirely to biometrics. This is a company that helped with the FBI’s WSQ fingerprint-compression work in the early era of digitized prints, then lived long enough to face a very different criminal instrument: a synthetic face generated on demand.
The scan is the easy part
The obvious way to describe Aware is as a maker of biometric algorithms. Its Nexa components match faces, fingerprints, irises and voices. AwareSDK, formerly the Knomi mobile framework, gives developers capture and authentication tools for phones. AwareID moves enrollment, authentication and document capture into a hosted service. AwareABIS searches large galleries for civil or criminal identification, while BioSP and Astra handle workflows, integration and the computing needed to search millions of records.
In practice, customers can assemble those pieces around the moment that matters to them. A bank can pair document capture with a face check for remote account opening, then ask for another biometric check when a transfer looks risky. An employer can replace a password at login. An airport or border program can search a traveler against a controlled gallery, while a police examiner can review candidates returned from a latent fingerprint. Developers get SDKs and APIs instead of building recognition science from scratch; administrators get workflows they can tune to policy. The same ingredients serve very different stakes, which is why Aware sells components as well as complete systems.
But a matching engine answers only one question: how closely does this sample resemble that template? A production identity system has to answer several more. Is the image usable? Is the person physically present? Should a weak match trigger another check? Is this customer allowed to keep biometric data in the cloud? What happens when a vendor performs well in one geography and poorly in another? The value shifts from recognition to decision-making.
This is the premise of the Awareness Platform, the company’s new center of gravity. Aware describes it as a SaaS control plane where an organization can design enrollment, verification, authentication and identification workflows; connect multiple biometric providers; normalize their outputs; and test one vendor against another without tearing apart a live deployment. Integrations announced with matching specialist ROC and identity-verification company Mitek make the positioning unusually explicit: Aware does not have to pretend its own component is best at every task if it can own the layer that chooses among them.
“No identity provider is number one at everything” is less a concession than Aware’s product thesis.A practical reading of the orchestration strategy
From lock-in to a mixing board
Large biometric deployments accumulate. A border agency may have one fingerprint matcher, another face engine, a document reader and rules written for several generations of hardware. A bank may use separate vendors for onboarding, account recovery and high-risk payments. Replacing the whole stack is dangerous; leaving it untouched is expensive. Orchestration offers a middle course: route a transaction through different checks, compare results and change the flow without rebuilding every customer experience.
Point solution
- One vendor, one check
- Output stays in its own format
- Switching means integration work
- Performance judged in isolation
Orchestrated workflow
- Several signals in one decision
- Outputs normalized
- Vendors tested in parallel
- Rules tuned for risk and friction
Aware’s own 2026 survey found that organizations already using biometrics ran three vendors on average, and 98 percent expressed interest in orchestration. Those numbers come from company-sponsored research, so they are evidence of a sales thesis rather than a neutral census. Still, the underlying problem is recognizable. Deepfakes, injection attacks and stolen identity documents have turned a neat login feature into a moving security program. A system that cannot change quickly is a system waiting for the fraudster to catch up.
Intelligent Liveness is Aware’s answer at the front of that contest. Launched in 2025, it combines passive analysis - checks the user need not notice - with active techniques that can ask for additional proof. The company says internal benchmarking cut false negatives by roughly half while keeping capture below a second. More important than a single metric is the update model: Aware says the machine-learning system can be retrained as new spoof methods arrive. In liveness detection, the finish line moves whenever a better face generator ships.
A specialist’s awkward economics
The strategic idea is cleaner than the income statement. Aware recorded $17.293 million in 2025 revenue, down slightly from $17.389 million in 2024, and lost $5.9 million. Maintenance supplied 51 percent of sales, software licenses 42 percent, and services and other revenue 7 percent. Research and development consumed $8.3 million - nearly half of revenue. Government buying cycles and large perpetual licenses make quarters lumpy, while the hosted subscription business is still developing.
How Aware made its 2025 revenue
This creates the central business challenge. Aware wants platform economics - recurring subscriptions, reusable workflows and a wider commercial market - while supporting customers that value long-lived, customer-controlled systems and careful procurement. The company sells through integrators, original equipment manufacturers and resellers because a matching algorithm is often one component in a national or enterprise system. That reach is useful, but it can put Aware one layer away from the final buyer.
The competition is broad. IDEMIA, NEC and Thales bring scale and government relationships. Entrust, Mitek, iProov, Jumio and Veriff compete around digital onboarding and fraud. ROC is both a specialist alternative and, now, an announced technology partner. Aware’s plausible distinction is not size. It is the combination of multimodal components, standards work, deployment history, customer-managed options and a willingness to sit above other vendors.
Trust is a systems property
Biometrics vendors like to speak about accuracy, but users experience consequences. A false rejection is a traveler stuck at a gate or a customer unable to reach an account. A false acceptance is a fraud loss. A stored template is not merely another password hash; a person cannot reset a face. The design choices around capture, storage, retention, review and fallback matter as much as the glamorous model in the middle.
Aware leans on certifications and external evaluation to make that argument credible. In 2025 it earned FIDO Alliance face-verification certification, reported ISO 27001 security certification and cited independent liveness and bias testing. It also highlighted its performance in the Department of Homeland Security’s Remote Identity Validation Rally. These marks do not settle every question about fairness, privacy or deployment. They make the system measurable, which is a more useful starting point than a promise.
CEO Ajay Amlani arrived in February 2025 with a biography unusually fitted to this moment. He co-founded the entity that became CLEAR, worked in government identity and defense innovation, and held leadership roles at IDEMIA and iProov. His task is not to persuade the world that biometrics exists; billions of phone unlocks have done that. It is to persuade cautious organizations that a small public company can govern an expanding collection of biometric decisions.
Aware’s long history helps and complicates the pitch. Legacy products provide expertise, customers and maintenance revenue. They also belong to an era when identity systems were purchased as monuments. The Awareness Platform imagines them more like a mixing board: signals enter, operators adjust the levels, and no single instrument has to play the whole song. If Aware’s bet works, the most consequential software in the biometric stack may be the part that knows when not to trust the face matcher.