Oscilar

Lenders lost $3.3B to people who don't exist in the first half of 2025... And that was before the current wave of AI-powered synthetic identity tooling really hit its stride. Most of this never even touches the fraud queue. These applications pass onboarding, clear underwriting, and get booked like any other loan. The loss only shows up later, on the credit side. Fraud and credit are still evaluated separately. Different systems, different data, different timing. By the time signals connect, the account is already on the books. Synthetic identities are built to take advantage of that gap. AI makes them faster to create, easier to scale, and harder to distinguish from real borrowers. This Thursday, we're getting into what's actually working in production for lenders: - How AI can enable real-time risk signals, more accurate approvals, and clearer, more explainable decisions - Best practices to reduce fraud, improve approval rates, and build a more adaptive risk framework. - How to unify risk data, automate decisioning, and reduce fraud without adding friction to the borrower experience. Feat. Saurabh Bajaj (Oscilar) I Tomás Campos (Spinwheel) I Phil Goldfeder (American Fintech Council) 🗓 Thursday, April 23 | 9:00 AM PT (this week!) 💻 Live webinar 🔗 https://bit.ly/4vCTfHE

Credit underwriting is breaking in the AI era. On the surface, everything looks like it works. Decisions get made. Models assign risk. Portfolios perform…until they don't. Fraud is now virtually indistinguishable from legitimate borrower activity. Regulators expect every decision to be replayable on demand. And the data landscape has expanded beyond bureau scores into signals most systems can't reconcile. AI raises the ceiling of what's possible. But layered onto fragmented architecture, it just inherits the same flaws. The institutions getting AI in credit right aren't just adding AI to their existing stack. They're fundamentally rethinking how decisions are made and governed at scale. We wrote a playbook on AI in credit underwriting, directly based on what we're seeing in production from early stage lenders to institutions serving millions. Inside: - Why misclassification (fraud booked as credit loss) is the failure mode most lenders miss - What regulatory convergence on decision traceability means for your stack - How $3.3B in synthetic identity exposure is passing traditional underwriting - What AI decisioning actually requires in production and how you can get up and running without breaking anything Preview linked in comments.

We're proud to share that Oscilar has been named to the Chartis Research FCC50 for 2026, with category awards in Agentic AI Innovation and Low-Code / No-Code Customization. Chartis evaluates hundreds of vendors globally based on real-world deployments. They don’t rely on self-reported data or paid placements. That methodology is what makes the FCC50 a practical shortlist for institutions evaluating production-ready systems. One thing that stands out this year is the introduction of the Agentic AI category. Most financial crime systems still stop at detection. They generate scores and alerts, and from there the work becomes manual. That gap between signal and action is where cost accumulates and where fast-moving fraud goes unresolved. Agentic systems close that gap by reasoning across context, investigating in real time, and taking action where appropriate. A heartfelt thanks to Chartis for the recognition. This honor belongs to the entire Oscilar team, and to the customers who chose to run their risk operations on a platform that didn't exist five years ago. Your trust is our lifeblood.

Fintech is at an AI crossroads. One camp is trying to replace humans. The other is trying to empower them. The "replace" vision sounds clean in theory: fully agentic systems, zero humans in decisioning, instant outcomes. Clean, fast, efficient. I understand the logic. But in financial services, that breaks down fast. Risk decisions carry real weight: A fraud flag that locks someone out of their account, an AML escalation that ends a banking relationship, a credit denial that impacts a business. These aren't just workflows. They require judgment. Accountability. And a human who can explain the decision to a regulator, an auditor, or the customer on the other end. That's why "replace the human" is the wrong goal. The real opportunity is leverage: AI that makes humans dramatically better, not irrelevant. Faster context. Clearer signals. Less mechanical work. Better decisions. That changes how you build: - AI that prepares the relevant cases, not closes them - Agents that surface the full picture before an analyst opens the file - Explainability as infrastructure, not a reporting afterthought This is what we're building at Oscilar. L1 agents reduce false positives so analysts focus on judgment. Organizations staying one step ahead of attacks due to rule recommendation agents. Every agent recommendation and disposition comes with a clear, human-readable audit trail. The result isn't slower because humans are in the loop. It's faster and defensible because they are 10x more productive. In a regulated industry, that distinction is everything.

Synthetic identity fraud surged 8x in 2025 (LexisNexis). One in nine fraud cases now involves a fake identity built from real data, costing U.S. lenders $3.3B in H1 alone (TransUnion). The signals were there, but most systems weren't designed to connect them. Join our own Saurabh Bajaj, Tomás Campos (Spinwheel), and Phil Goldfeder (American Fintech Council) for practical look at how leading lenders are using data-rich AI to unify risk data and make smarter decisions, without adding friction: → See the full borrower picture across fragmented data, in real time → Approve more borrowers accurately, without increasing exposure → Detect sophisticated fraud (including synthetic identities) that rules-based systems miss 🗓 Thursday, April 23 | 9:00 AM PT 💻 Live webinar 🔗 Link to register in comments

Heads up: Proposed changes to AML/CFT just shipped. ⚠️ This week, the FDIC, OCC, NCUA, and FinCEN moved together to overhaul AML/CFT requirements across U.S. banks and credit unions, impacting roughly 8,100 institutions. The coordination is deliberate. The Bank Secrecy Act (BSA) has anchored AML since 1970, but it was built for a world of paper trails, batch processing, and correspondent banking. That world no longer exists. This regulatory posture affirms that. Three things stand out: → Risk-based monitoring is now mandatory. Blanket controls are out. Resources must align to risk: more for high-risk customers, less for low-risk. → "Effective" has a new meaning. Programs must work in practice and hold up under scrutiny. Gaps in monitoring, data, or risk assessment will draw continuous supervisory attention. → FinCEN's role just expanded. A new consultation layer gives FinCEN a formal role before enforcement actions move forward. FDIC Chair Travis Hill explained that the goal is to stop penalizing banks for "foot faults" and "box-checking exercises" and start focusing on whether institutions are actually identifying and mitigating risk. That completely changes the core compliance question. Did your program work, or did it just run? Most legacy AML infrastructure wasn't built to answer that. At Oscilar, we've always operated on the assumption that process completion alone wouldn't be enough. Now regulation is catching up. The official comment period is now open. #FRAML

We're headed to Charlotte next week for Fraud Fight Club! 💥 🥊 Saurabh Bajaj and the Oscilar crew will be there, come say hello at Booth #2. We've also got a great side event with SentiLink right after Day 1. An evening of networking, food, drinks, and a little friendly competition at Tipsy Pickle, just steps from The Revelry. Think golf simulator, pool tables, darts, and a relaxed space to connect with fellow fraud fighters. 🗓 Tuesday, April 14 I 6-8 PM 📍 Tipsy Pickle 🔗 Space is limited, link to request a spot in the comments See you there!

🚀 Oscilar na FEBRABAN SEC 2026 🇧🇷 Nos dias 18 e 19 de março, estaremos na FEBRABAN SEC em São Paulo para mostrar como estamos devolvendo o sistema operacional de risco AI-native para casos de uso como: #Onboarding de clientes PF e PJ - Concessão de #crédito #credit - Prevenção a #Fraude #fraud - Prevenção a Lavagem de Dinheiro #pld #aml #tm A Oscilar é uma #plataforma #AI Native e low-Code que permite que times de Fraude, Crédito, Risco e Compliance operem e atualizem suas políticas de risco em tempo real de forma autônoma e com apoio dos nossos #agentes #agents. Chega de depender de ciclos e sprints de #engenharia e #tecnologia #tech para proteger sua operação. O que você vai encontrar: - Autonomia Total: Crie, teste e coloque políticas de decisão em produção em minutos, sem precisar de código; - IA que Toma Ação: Conheça nossos AI Agents que automatizam investigações complexas e eliminam backlogs de alerta; - Decisões Unificadas: Onboarding, Crédito, Fraude e AML em uma única camada de dados inteligente; - Redução de Falsos Positivos: Melhore a experiência do cliente aprovando bons usuários sem fricção desnecessária. 📍 Estande 15 📅 18 e 19 de Março Não deixe sua estratégia de risco para o próximo ciclo orçamentário. O futuro da decisão é agora. Nos vemos lá! Lourenço Biselli Steve Parke Sachin Kulkarni Saurabh Bajaj Neha Narkhede Linas Beliūnas Tomas Lobo Lua Bustamante Alesia Do Isabella Ribeiro Max Del Valle Derek Carranco Jessica Jönzzon (Whitney) Dominic Glover

Jensen Huang, Jeff Bezos, Sam Altman, Larry Fink, Taylor Swift, Neha Narkhede. What do these people have in common? They've been named to the Forbes 250: America's Greatest Innovators. And our Co-Founder and CEO is one of the few on the list currently building her second company from the ground up. Forbes 250 recognizes the people who shaped the currents of the modern world: the software that made ".com" possible, the math that makes the internet work, the invention of 3D printing, the gene editing technology that's rewriting the future of medicine, the electronic spreadsheet. Neha is recognized for architecting the data infrastructure that makes an AI-enabled internet possible. She co-created Apache Kafka, the real-time data streaming standard used by over 90% of the Fortune 100. She co-founded Confluent, taking the ethos of Kafka from an open-source project to a public company, recently acquired by IBM for $11B. After nearly two decades of building at that scale, she had certainly earned the right to step back. But she didn't stop. She started over. Neha has always been drawn to the infrastructure problems that everyone else routes around. At LinkedIn, it was real-time data streaming. At Confluent, it was making that infrastructure accessible to every enterprise. At Oscilar, it's the gap between real-time data and real-time risk decisions. It's the same instinct that has defined her career: find the infrastructure problem an industry is working around and build the platform that solves it directly. In her own words this week, Neha reflected on her journey: starting out in rooms where she was often the only one who looked like her, earning her place over and over, keeping her head down and working harder than she should have had to. Over time, that changed. She started speaking up, creating space for others, and thinking about the kind of environment she was building, not just the product. That's something we live every day at Oscilar. Neha is deeply involved in every aspect of how our company operates — in product decisions, in how we work with customers, and in the culture we're building. She sits in design reviews. She pushes back on anything that feels like a shortcut. When customers like SoFi ask for something, it shapes what we actually build next, not just what goes on a roadmap. Oscilar is trusted by institutions managing trillions in assets and the fastest-growing players across four continents. They understand that risk isn't a checkpoint. It's a living, connected, continuous system. Every decision is stronger when it carries the full context of the customer journey, in real time. That's why institutions like MoneyGram, Nuvei, Payoneer, and Uphold trust Oscilar to move at the speed of the transaction with connected, full-context, AI-native intelligence. Forbes recognized an innovator building foundational technologies. We get to work with her every day. Just wait until you see what comes next. Congratulations, Neha.