About RTScale

The consent-provenance layer
for moments that can't be undone.

RTScale started with a narrow question: what happens to consent when the window between intent and irrevocable settlement collapses to zero?

The payment rails got faster. FedNow, RTP, instant card auth — settlement windows that once stretched days now settle in seconds. The fraud infrastructure didn't keep pace. Device fingerprinting and behavioral biometrics can tell you whether the right account is being used. They cannot tell you whether the person using it is acting freely.

Authorized-push-payment scams, elder financial exploitation, coerced wire transfers — these attacks pass every conventional check because the legitimate account holder is the one pressing approve. The account is right. The device is right. The consent is the question.

The answer we kept returning to was an artifact: a compact, cryptographically signed record of observable affective state at the moment of action. Hardware-rooted so it cannot be spoofed by a replay attack. Decomposed so it carries affective indicators — not facial-recognition embeddings, not voiceprints, not raw biometric data. Portable across custody chains, regulatory regimes, and dispute resolution processes. Erasable in a way that satisfies GDPR Article 17 without destroying the evidentiary chain.

We called it the State of Mind Signature. We built it first for banking and payments — the category where the problem is most acute and the regulatory pressure most immediate. Then we asked whether the primitive generalized. It did. Wire fraud. Elder property exploitation. Agentic AI systems authorizing financial actions on behalf of principals. Self-custody crypto wallets under physical coercion. Every case fit the same pattern: a high-stakes, hard-to-reverse moment where the existing evidence of consent was a signed document, a checkbox, or an OAuth token — none of which document the state of mind in which the consent was given.

RTScale is infrastructure for that gap. Not a fraud detection tool. Not a compliance checkbox. The consent-provenance layer for any consequential transaction that is hard or impossible to reverse — at the moment the consent is given, where the existing evidence is thinnest.

Mission

RTScale builds the consent-provenance layer for transactions that are hard or impossible to reverse — at the moment the consent is given, where the existing evidence is thinnest.

Values

How we work and why.

  • Consent over compliance.

    A signed document proves someone acted. It does not prove they understood what they were signing, or that the action was free. We build for the gap between the signature and the consent — not because regulators require it, but because it's the honest measure of what 'authorized' should mean.

  • Precision over persuasion.

    We say what the technology does and what it does not do. The SoM Signature captures observable affective indicators — not inner emotional states, not clinical diagnoses, not moral judgments. Overstating what an ML model can determine is a fast path to misuse. We do not take that path.

  • Privacy by architecture.

    Raw camera frames and audio buffers stay on the device. The cloud receives a decomposed-indicator artifact — affect vectors and keyword-bound microexpression events, not facial-recognition embeddings or voiceprints. Cryptographic erasure satisfies right-to-erasure without deleting the evidentiary record. These are not privacy features added on top; they are the architecture.

  • Trust earned through disclosure.

    Our trust center is not a checkbox. It publishes what we collect, what we do not collect, what our subprocessors touch, and what our responsible-AI framework rules out entirely. Security review is invited. Scrutiny is load-bearing for a product that asks partners to put us in the consent chain of consequential transactions.

  • Irreversibility as the design constraint.

    We started with a narrow question: what happens when the window between intent and irrevocable settlement collapses to zero? The answers turned out to be structural — not just fraud, not just elder exploitation, not just coerced transfers, but a category. Hard-to-reverse moments are underserved by the existing consent infrastructure. That is the category we are building.

Team

Leadership

Peter Walker headshot

Peter Walker

Co-founder & CEO

The Architect of High-Performance Systems

Dr. Peter Walker is a rare "dual-threat" leader who bridges the gap between deep-tech invention and Fortune 50 executive strategy. A pioneer in low-latency distributed systems, Peter's career is defined by solving engineering problems others deemed impossible. He first proved this by founding Spawn Labs, where he architected the industry's first cloud gaming platform ("Netflix for games"), leading to a successful acquisition by GameStop in 2011. Following this exit, Peter transitioned to global scale, serving as VP of Architecture at Visa, where he spearheaded the "AI@Scale" initiative to integrate deep learning into the world's largest payment pipeline. Most recently, as Corporate Vice President at Microsoft (Industry Solutions Engineering), he led the engineering teams solving critical technical challenges for the world's largest financial institutions. At RTScaleAI, Peter is applying his multidisciplinary expertise—spanning AI, real-time streaming, and behavioral science—to solve the rapidly growing multi-billion dollar problem of authorized payment fraud.

Vivek Iyer headshot

Vivek Iyer

Co-founder & CTO

The Technical Moat & Patent Engine

Vivek Iyer brings the technical defensibility that defines a true Deep Tech seed investment. As CTO, he is responsible for RTScale's intellectual property strategy and the "Trust Cortex" architecture. Vivek is not just an engineer; he is a prolific inventor with over 100 granted and pending U.S. patents spanning AI, real-time systems, and behavioral analytics. His technical authority is undisputed: at Dell Technologies, he served as a Senior Distinguished Technologist and was named "Inventor of the Year" for four consecutive years (FY2019–FY2022)—an unprecedented feat in a company of 100,000+ employees. Prior to Dell, Vivek served as Director of Engineering at Spawn Labs (acquired by GameStop), where he built the ultra-low-latency streaming infrastructure that underpins modern cloud gaming. Vivek left a distinguished corporate career to co-found RTScaleAI because he identified "Intent Verification" as the next massive shift in real-time computing.

Matt Walker headshot

Matt Walker

Co-founder & COO

The Operational Architect of Behavioral Trust

Matt Walker is an operational strategist who bridges the gap between abstract R&D and commercial scalability. As Co-Founder and COO, he is the architect behind RTScale's "Behavioral Trust Layer" concept, translating complex "State of Mind" metrics into audit-ready, compliance-friendly banking solutions. Unlike typical founders focused solely on product, Matt's expertise lies in constructing the operational backbone required to deploy high-risk AI technologies into regulated markets. His reputation for operational excellence was cemented during an 8-year tenure as Director of Operations at Vinli (Series B connected-car platform), where he engineered a full-scale transition to remote work that drove 68% growth and maintained 92% staff retention. With a background spanning Private Equity due diligence (Bridgewater Associates) and regulatory compliance (GDPR, ISO, SOC), Matt provides the "fixer" DNA that investors seek—a leader who ensures that technical innovation survives the rigors of bank procurement, audit, and global regulation.

Subir Roy

Co-founder & VP of Engineering

The Enterprise Scaler

Subir Roy provides the "bank-grade" reliability assurance that RTScaleAI needs to sell into financial institutions. He joins the team from Visa, where he served as Director of the Data and AI Platform, engineering real-time machine learning infrastructure that processed over 1 billion transactions daily with 99.999% availability. Subir's expertise is in "AI under pressure." He built Visa's Real-Time AI Platform from inception, scaling it to handle 65,000 transactions per second with sub-10ms latency. His background also includes pivotal roles at Ixia (acquired by Keysight for $1.6B) and Cisco, where he developed award-winning network intelligence products. At RTScaleAI, Subir ensures that the "State of Mind" models don't just work in the lab—they work at the speed of the global payments rail.

Oswald Becca headshot

Oswald Becca

Sr. Platform Engineer

The Latency Specialist

Oswald Becca completes the "Spawn Labs Reunion" team, bringing over 20 years of hardware and systems validation experience to the platform. A co-founder of Spawn Labs alongside Peter Walker, Oswald architected the remote console emulation that led to the company's acquisition by GameStop. Currently an Emulation Engineer at Intel, Oswald specializes in the "hard tech" of system validation and zero-latency performance. His ability to bridge silicon-level constraints with high-level software requirements provides RTScaleAI with a unique advantage in optimizing the "Identity Shield" for edge devices and mobile SDKs.

Work with us. Evaluate us. Challenge us.

Design-partner conversations are open. We're building in public where we can — trust center, protocol specs, compliance library.

Schedule a working session Trust center