Taktile

Most lenders manage risk through upfront checks. Product leader Peter Lu shares how his team flipped the script, using product design itself as the risk control. Before joining Monzo, Peter's team had to launch a credit product in a strict market by issuing real, unsecured lines to users whose credit profiles ranged from subprime to credit-recovering. Traditional underwriting alone wouldn’t work. Here's how Peter's team approached it: → 𝐓𝐡𝐞𝐲 𝐫𝐞𝐟𝐫𝐚𝐦𝐞𝐝 𝐭𝐡𝐞 𝐩𝐫𝐨𝐝𝐮𝐜𝐭, 𝐧𝐨𝐭 𝐭𝐡𝐞 𝐫𝐢𝐬𝐤 𝐦𝐨𝐝𝐞𝐥: Instead of positioning it as a credit card, they positioned it as a credit-builder that will actually support customer goals. → 𝐓𝐡𝐞 𝐟𝐫𝐚𝐦𝐢𝐧𝐠 𝐜𝐡𝐚𝐧𝐠𝐞𝐝 𝐮𝐬𝐞𝐫 𝐛𝐞𝐡𝐚𝐯𝐢𝐨𝐫: Customers who came in to build credit had little reason to draw on the line, which kept utilization (and risk) naturally low. → 𝐓𝐡𝐞 𝐫𝐞𝐬𝐮𝐥𝐭: ~$250 in credit per user, no traditional underwriting checks, default rates under 5%, and a margin-positive product. For more on Peter’s approach to product and risk, check out the full conversation on our blog 👇

What’s Taktile’s case for using agentic AI in commercial underwriting? We're seeing credit teams expand automation without losing rigor. Here’s how we built that trust: → 𝐔𝐧𝐝𝐞𝐫𝐬𝐭𝐚𝐧𝐝 𝐰𝐡𝐚𝐭 𝐦𝐚𝐤𝐞𝐬 𝐜𝐨𝐦𝐦𝐞𝐫𝐜𝐢𝐚𝐥 𝐮𝐧𝐝𝐞𝐫𝐰𝐫𝐢𝐭𝐢𝐧𝐠 𝐬𝐨 𝐜𝐨𝐦𝐩𝐥𝐞𝐱 Business documents vary widely, financial statements arrive in different formats, and important risk signals are often spread across multiple systems. → 𝐀𝐮𝐭𝐨𝐦𝐚𝐭𝐞 𝐭𝐡𝐞 𝐦𝐚𝐧𝐮𝐚𝐥 𝐰𝐨𝐫𝐤 𝐢𝐧 𝐜𝐚𝐬𝐞 𝐩𝐫𝐞𝐩𝐚𝐫𝐚𝐭𝐢𝐨𝐧 AI agents are well suited to the parts of underwriting that are document-heavy, multi-step, and context-dependent, making them reliable counterparts for speeding up the time to decision. → 𝐇𝐞𝐥𝐩 𝐭𝐞𝐚𝐦𝐬 𝐝𝐞𝐥𝐢𝐯𝐞𝐫 𝐟𝐚𝐬𝐭𝐞𝐫 𝐜𝐮𝐬𝐭𝐨𝐦𝐞𝐫 𝐝𝐞𝐜𝐢𝐬𝐢𝐨𝐧𝐬 𝐰𝐡𝐢𝐥𝐞 𝐦𝐚𝐢𝐧𝐭𝐚𝐢𝐧𝐢𝐧𝐠 𝐫𝐢𝐠𝐨𝐫 With richer context from agentic AI and more time to focus on what needs their expertise, credit teams can enhance the way they make underwriting decisions. To learn about real-world applications of AI in commercial underwriting, check out our blog at the link in comments 👇

Incredible to see Sahasranaman Ranganathan and the Made Card team bringing a sharper underwriting edge to consumer credit, with a great breakdown of the strategy behind it. This is how our customers are pushing the frontier of credit and risk operations. Check it out 👇

Our Iași team grew from 3 to 13 engineers this year. That's 333.33% more people reshaping financial services with AI (and 1,200% more screens). 👾 👾 If you want to be part of what we’re building, check out our open roles at the link in the comments 👇

AI is moving fast in financial services. Our newsletter helps teams stay ahead. Each edition of AI Field Notes brings you: → Conversations with leaders across credit, fraud, AML, and more → Practical techniques from engineers building AI workflows in production → What's working across the industry Check out this month's edition, or subscribe to get the next one in your inbox (linked in the comments).

Synthetic identity fraud is no longer an edge case, and AI is making it harder to detect by the day. Our reseller partnership with SentiLink is built to help teams meet the moment. SentiLink is a leading identity fraud detection platform, trusted by 13 of the top 15 U.S. banks, six of the top 10 credit unions, and 50+ fintech unicorns to stop synthetic identities, stolen identities, and first-party fraud at the point of application. With SentiLink’s products directly available through Taktile, fraud and credit teams can embed specialized identity intelligence into their decision flows, without lengthy procurement or engineering work. Together, we're helping financial institutions catch sophisticated fakes earlier while approving more legitimate thin-file customers with confidence. Learn more on our blog. Linked in the comments below ⬇️

Taktile’s sliding into summer in NYC with a Yankees baseball game ⚾️ And yes, we came with the merch to match!

Product and risk are often treated as separate functions with separate goals. Monzo Bank's Peter Lu argues product and risk are intertwined, and that treating them as one starts with how leadership sets the strategy. Take a fintech entering the US expat market. Peter explains how product and risk would work as one from the start: → 𝐒𝐭𝐫𝐚𝐭𝐞𝐠𝐲 𝐚𝐥𝐢𝐠𝐧𝐬 𝐟𝐫𝐨𝐦 𝐭𝐡𝐞 𝐭𝐨𝐩: Leadership sets the ICP, and product and risk build their plans against the same definition of who the customer is. → 𝐑𝐢𝐬𝐤 𝐡𝐞𝐥𝐩𝐬 𝐢𝐝𝐞𝐧𝐭𝐢𝐟𝐲 𝐭𝐡𝐞 𝐝𝐚𝐭𝐚 𝐫𝐞𝐚𝐥𝐢𝐭𝐲: Many expats arrive with no US credit score or digital footprint, making traditional underwriting signals hard to apply. → 𝐏𝐫𝐨𝐝𝐮𝐜𝐭 𝐝𝐞𝐬𝐢𝐠𝐧𝐬 𝐚𝐫𝐨𝐮𝐧𝐝 𝐭𝐡𝐞 𝐠𝐚𝐩: Alternative data sources, different verification flows, and signup experiences should be uniquely designed for the newcomer segment. For more, check out the full conversation on our blog (linked in comments) 👇

Few things beat a room full of fintech operators, founders, and friends catching up over cocktails. Thank you to everyone who joined our NYC happy hour last week, and to Baselayer Fundbox and Fintech NerdCon for co-hosting with us. Until next time 🥂

We heard our community. You want to up the ante on AI deployment, but using agents responsibly is top of mind. So, we gathered a bench of experts in AI, credit, AML, and compliance to create the resource you need to move forward with confidence. A Practical Guide to Responsibly Deploying Agents in Financial Services brings together actionable advice from Dustin J. Eaton CFE CAMS CFCI CFCS CAFP CGSS CAMS-RM CAFS, Maximilian Eber, and Jill Zucker Sheckman, It’s structured in three phases of learning designed to help you build an informed AI roadmap: → Identify your opportunities: Develop your understanding of the problems agents can solve, and zero in on the highest-impact areas to deploy them within your org. → Understand the infrastructure: Learn how to orchestrate agents for responsible governance and effective performance in production. → Activate your roadmap: Establish the step-by-step pathway to deploying your first agents, scaling them across your organization, and maintaining them over time. Read the guide via the link in comments 👇