Today, we’re excited to announce @Evervault's $25M Series B, led by Ribbit Capital with continued support from @sequoia, @IndexVentures, @kleinerperkins, and @nextplayVC. This round comes at a time when sensitive data exchange on the web is going parabolic. Since 2019, we’ve been focused on building durable infrastructure for engineering teams to collect, process, share, and enrich sensitive data -- while keeping it encrypted at all times. We thought we were making good progress in encrypting the web, helping customers like @tryramp, @Rippling, @finix, @TheOverwolf, @Uniswap, @CarTrawler, and hundreds of others secure more than $5bn/year in payment flows and 100m+ unique tokens per month. But the past year has shown that our enemy -- plaintext data -- is getting stronger and more pervasive. Our vision is to build the clearinghouse for sensitive data, helping companies exchange sensitive data in a secure and encrypted way. This round helps us encrypt more of the web by further refining our developer experience, building deeper integrations with trusted third-parties, and increasing the value we can offer our customers for more data types. First and foremost, thank you to our customers. You trusted Evervault to sit directly in the flow of your most sensitive data (payments, identity, financial information, and more) and that trust is not something we take lightly. Your feedback, your requirements, and the problems you bring to us every day are what shape the product and push us forward. Thank you to the Evervault team. What you’ve built is genuinely special: infrastructure that lets developers process sensitive data without ever having to see it in plaintext. The pace, craft, care, and ambition you bring to work every day are what makes this company what it is. And thank you to our new investors for believing in the vision of making security architectural rather than procedural. We’re grateful to have partners who understand both the scale of the problem and the opportunity ahead. The internet still assumes that sensitive data must exist in plaintext somewhere. We’re building the infrastructure to change that. Onwards! More here 👉 evervault.com/blog/series-b?…

Dots agents from @OpenAI can buy things for you, from stores where you've already saved a card (so the agent only shops where you already shop) The problem though (with this, and a lot of “agentic commerce”) are that the errands worth handing off to an agent are at stores that have never seen your card (or maybe that you’ve never even heard of). Single use payment credentials -- bound to a specific merchant <> cart <> consent combo -- are the approach likely to win out long-term

A Brazilian debit card starting 6505 came back from our BIN Lookup API as "Discover Credit". Even though Discover holds the whole 65xxxx range, Elo has ranges inside it and runs on Discover's rails abroad. The first six digits can't tell you which one you're processing. Getting accurate BIN data is 𝗵𝗮𝗿𝗱

On Wednesday, a customer asked if our card collection iframe could support WebMCP. Their ideal timeline was the end of October. We shipped it on Thursday. Evervault Card Collection now has an opt-in agentTools option. It registers three WebMCP tools inside the card iframe, so browser agents can fill in card details through the same validation and encryption flow as a person: - get-form-status: which fields are filled, valid or erroring, and whether the form is complete - focus-field: moves focus to a field - set-field-value: enters a value, validates it and returns the updated status Each tool maps to something a user can already do, and nothing more. Card data only leaves the iframe encrypted, and never in a tool result. The tools are only exposed to origins you allowlist, and white-label customers can name them after their own product. We've been seeing an explosion in agent-driven commerce through our platform in recent months, and are excited to be releasing more functionality to help solve the trust, security and consent challenges that appear when agents buy things for us.