Legacy finance is accidentally built for AI (8 minute read)
Finance's "outdated" infrastructure - COBOL, SWIFT/ISO/NACHA flat files, batch jobs - is actually an ideal substrate for AI. Money already moves as structured text, so AI agents can parse, translate, monitor, and augment legacy systems without ripping them out, turning digital transformation into AI "wrapping" rather than replacement. With tools from AWS, Microsoft, Oracle, and startups parsing ISO 8583 and mainframe code, AI augmentation may succeed where cloud-first modernization failed. |
a16z's David George on how private and public markets fused into one (5 minute read)
Andreessen Horowitz partner David George explains why today's largest tech companies feel no urgency to IPO: private capital markets have become deep and liquid enough that firms can raise funds and provide liquidity without ever going public. He also discusses structural challenges in public markets, employee retention tools in private firms, and how AI and market dynamics are shaping decisions about if and when private giants like OpenAI or SpaceX might finally list. |
How Zelle scaled to $1.2 trillion by embedding into core banking (7 hour read)
Zelle has quietly become a backbone of digital payments in the US, moving $1.2 trillion across 4.2 billion transactions in 2025. The network now reaches institutions representing 80% of US deposit accounts, with nearly a third of volume tied to small businesses and new cross-border ambitions emerging via stablecoin initiatives. For fintechs, integrating with Zelle offers instant access to 2,300+ financial institutions and real-time liquidity, but requires accepting compliance overhead and the risk trade-offs of irreversible payments in a fraud-sensitive environment. |
|
Kalshi and the rise of macro markets (40 minute read)
The Federal Reserve Board analyzes how Kalshi's regulated prediction markets generate real-time forecasts for macro variables and policy outcomes. Comparing Kalshi-implied expectations with surveys and traditional market measures, the authors find the platform provides high-frequency, well-calibrated density forecasts that respond meaningfully to economic news. The paper argues that these markets can serve as a complementary benchmark for measuring expectations and uncertainty. |
The road map for embedded insurance (5 minute read)
Embedded insurance is evolving from a simple distribution tactic into a growth engine that expands access across life, home, auto, cyber, and gig coverage. Yuri Poletto highlights how Brazil's NuBank sold 2 million life policies, illustrating how fintech platforms can unlock entirely new insurance demand rather than just shift market share. With savings of up to 40% in some commercial lines and rising adoption across proptech, telcos, auto manufacturers, and SaaS platforms, embedded models are increasingly reshaping how insurers, fintechs, and regulated partners split ownership, margins, and customer experience. |
|
Minions: Stripe's one-shot, end-to-end coding agents—Part 2 (8 minute read)
Stripe details the infrastructure behind "minions," its unattended coding agents that now produce 1,300+ merged PRs per week. Minions run in pre-warmed, isolated cloud devboxes and are orchestrated via "blueprints"—a hybrid of deterministic workflow steps (linting, pushing, CI hooks) and flexible agent loops (implement task, fix CI). With scoped rule files, a centralized MCP tool server ("Toolshed") offering ~500 tools, and tight CI feedback limits, Stripe optimizes for one-shot, end-to-end PRs that are fully agent-written and human-reviewed. |
CME group to launch 24/7 crypto futures trading (3 minute read)
CME Group is preparing to offer around-the-clock cryptocurrency futures trading on its CME Globex platform starting May 29, pending regulatory approval. The move comes as digital asset derivatives activity surges, with 2025 notional volume hitting $3 trillion and average daily volume in 2026 climbing 46% year over year to 407,200 contracts, driven largely by Bitcoin and Ethereum futures. By aligning crypto trading hours with the always-on nature of digital markets, CME is tightening the link between traditional finance infrastructure and institutional crypto demand. |
|
Klarna pursues more banking customers (4 minute read)
Klarna is leaning harder into debit cards and longer-term lending as it works to offset mounting quarterly losses and deepen relationships beyond buy now, pay later. The company reported a $26 million fourth-quarter loss despite revenue climbing 38.5% year over year to $1.08 billion, with CEO Sebastian Siemiatkowski arguing that surging loan originations will pressure margins now but drive higher lifetime profits over time. With 118 million active users globally, Klarna is betting that turning frequent checkout users into full-fledged banking customers will boost retention, expand margins, and stabilize its path to profitability. |
From a mall booth to a $3B fintech: Matt Oppenheimer's startup lessons (6 minute read)
After nearly 15 years as CEO, Remitly co-founder Matt Oppenheimer is stepping into a chairman role, reflecting on lessons from scaling a three-person Techstars startup into a public remittance company with $442M in Q4 revenue and its first full year of GAAP profitability in 2025. He emphasizes falling in love with the problem (not the product), staying obsessively close to customers, defining culture through behaviors, raising sufficient capital, and treating leadership growth as an intentional, evolving process. |
SaaS markets have crashed in 2026, but is private credit the even bigger risk? (12 minute read)
Jason Lemkin argues the scarier story than falling SaaS multiples is the private credit system behind software buyouts: roughly $600–750b of exposure could get stressed as AI-driven seat compression and growth slowdowns hit cash flows. He points to rising distressed tech debt, "hung" loan deals, misclassified software exposure inside BDC portfolios, and a 2026 maturity wall as ingredients for a slow-moving credit squeeze that tightens refinancing, M&A, and venture debt. The piece lays out a potential doom loop where delayed marks meet redemptions in illiquid vehicles, forcing fire sales and spreading pressure beyond just software equities. |
Infosys & Anthropic AI in manufacturing, telco, & finance (6 minute read)
Infosys is teaming up with Anthropic to bring enterprise-grade, agent-driven AI into telecom, financial services, manufacturing, and software development, combining Claude models with its Topaz platform. The partnership aims to close the gap between flashy AI demos and systems that can reliably operate in highly regulated industries, starting with telecom before expanding into finance and manufacturing. In financial services, the focus includes risk detection, automated compliance reporting, and more personalized client interactions, positioning agentic AI as a practical tool for modernizing core operations rather than just boosting productivity. |
|
| Love TLDR? Tell your friends and get rewards! | | Share your referral link below with friends to get free TLDR swag! | | | |
| Track your referrals here. |
|
| |
0 Comments