Napier AI

Great session on Wednesday at the FinCrime Leaders Summit East Coast in Charlotte. It was an honor to join Thomas J. Nadratowski (Citi), Sean Meisler (USAA), and Ervin Brabham (AWS) on stage to discuss the operational reality of AI adoption and model governance. The conversation moved beyond the theoretical and straight into how we are fundamentally redesigning financial crime operating models in light of the recent FinCEN and SR 26-2 guidance. Here are my three key takeaways from our discussion: Target the "Noise" First: Find the lowest risk and most repetitive manual tasks in the investigative workflow. In my opinion, the highest impact right now is getting rid of the noise by automating the L1 alert review process.. By using AI agents to adjudicate low-risk false positives in milliseconds, firms are seeing 60-90% increases in operational efficiency, allowing investigators to focus on actual risk. Layer, Don’t rip and replace: You don't need a multi-year "rip and replace" program to see results. By layering AI into legacy infrastructure, financial institutions can move from implementation to production in weeks, not months—often with zero IT resources required. The New Governance Standard: With the shift from SR 11-7 to SR 26-2, regulators are moving away from check-the-box compliance toward risk-based judgment. Vendors must provide full decision architecture—traceable data lineage and sensitivity analysis—that proves to internal auditors and regulators exactly how accuracy and coverage are achieved. The industry is moving from simply making existing processes "faster" to making them fundamentally "different." Excited to keep pushing the needle on what's possible in the AI compliance space! Thanks to Transform Finance #TransformFinance for a well executed event. If you missed this panel and want to learn more about AI and SR 26-2, Castellum.AI is hosting a fireside chat on Tuesday talking about SR 26-2 and Model Risk Management for Agentic AML Systems - Register here https://lnkd.in/ehfTprJe

The single most important structural decision in ABM is account tiering. There are different approaches to tiering accounts, but here’s an example: 1️⃣ One-to-One (Tier 1): Top 10 accounts per market. Maximum personalisation. Custom content, direct mail, dedicated executive engagement, bespoke event invitations, and sales fully integrated into every touch. 2️⃣ One-to-Few/Many (Tier 2): Next 20 or remaining accounts per market. Segment-level personalisation. Industry-specific messaging, thought leadership, tailored case studies, targeted advertising, curated event invites. The discipline of tiering is saying no to campaigns, to events, to content that doesn't serve your highest-value accounts first. 💡 Hear it from: Jarod Horan , Strategic Key Account Director, APAC, Anteriad : Programmatic display is one of the most under-valued tools in an account tiering strategy (and one of the most misunderstood). At its core, it lets you follow your target audience around the web, building a consistent brand presence across all tiers to improve recognition, engagement, and recall in a way that most single-channel plays can't replicate. Here's where tiering and programmatic really intersect: Reserve your richest formats - interactive video, rich media - for the accounts that matter most. These formats go beyond awareness. They answer the question your audience is quietly asking: why should I care about this brand? This matters more than most marketers realise. The feedback I consistently hear in market is that audiences often recognise a brand name but have no idea what the company actually does. Rich, interactive formats close that gap. They build brand affiliation and understanding at the same time. Programmatic isn't just reach. Applied to your account tiers, it's one of the most efficient ways to make your entire ABM programme work harder. — What tier model does your team currently use and where does programmatic fit in your mix? Would love to hear how others are structuring this. #b2bmarketing #accountbasedmarketing #abm