Mercury CEO: three years of profitability, accelerated FDIC insurance, and AI cutting 40% of support tickets
Nov 7, 2025 with Immad Akhund
Key Points
- Mercury reaches three consecutive years of profitability, positioning sustained earnings as proof of durability to customers amid volatile funding cycles.
- An AI chatbot resolves 40% of Mercury's incoming support tickets, replacing a legacy suggested-topics system with direct answers that drive higher customer engagement.
- Mercury's CEO has formalized a $26 million angel fund after a decade of roughly 350 investments, with portfolio heavily concentrated in vertical B2B AI applications.
Summary
Mercury has now logged three consecutive years of profitability, a milestone the company disclosed on November 7 via Fortune's Term Sheet. The neobank, which targets venture-backed startups and small businesses, frames sustained profitability not as a financial luxury but as a core product signal — evidence to customers that the platform will remain operational regardless of funding cycles.
Trust as Infrastructure
The SVB collapse in 2023 crystallised Mercury's positioning. SVB carried roughly 80% uninsured deposits at the time of its failure, leaving customers dangerously exposed. Mercury responded by building out accelerated FDIC insurance coverage, publishing a floor of $5 million per account on its website, though the actual coverage extended to most accounts significantly exceeds that figure. According to Mercury's CEO, nearly all deposits held at Mercury are now FDIC insured — a structural differentiation from the concentrated, uninsured deposit base that brought down SVB.
Mercury has grown from a standing start in 2019 — when the CEO noted the idea of a modern business bank "sounded weird" — to holding accounts for customers with balances exceeding $100 million. The company now serves over 200,000 business customers and has expanded well beyond core banking into credit cards, bill pay, and invoicing.
AI Driving 40% Reduction in Support Tickets
AI deployment at Mercury is concentrated in two areas. On the back office side, the company is using AI to automate or augment compliance and document review workflows — formation documents, KYC checks, and similar processes that previously required significant human labour. The CEO frames this as a cost-saving exercise rather than a revenue driver.
On the customer-facing side, an AI chatbot now resolves 40% of incoming support tickets, handling routine queries such as wire settlement timelines. The CEO clarified this replaced the previous live-messaging suggested-topics system rather than the legacy email channel, noting that customers engage far more readily with a direct AI answer than with traditional FAQ suggestions.
Inference costs are not yet a material margin concern. Mercury's relatively contained customer base of 200,000 accounts, combined with a utility-driven usage pattern rather than high-frequency consumer engagement, keeps per-customer inference spend modest. The CEO acknowledges costs will rise as AI is applied to transaction processing at scale but expects it to remain a small percentage of revenue given Mercury's higher per-customer economics compared to prosumer SaaS peers.
Angel Portfolio and AI Investment Thesis
Mercury's CEO has made approximately 350 angel investments over ten years and earlier in 2025 formalised that activity with a $26 million angel fund. The current portfolio is heavily AI-oriented. The CEO sees the early wave of AI dev-tool investments as largely played out, with capital now flowing into deeper vertical B2B AI applications.
His broader thesis is that AI unlocks labour categories — physical through robotics, intellectual through software — that were previously unattractive to venture. Services businesses, long dismissed by institutional investors, are now investable as AI-plus-human hybrid firms with a plausible path to full automation. He draws a direct parallel to Marc Andreessen's "software is eating the world" framing, positioning AI as eating the remaining edges.
On SaaS and the Multiproduct Shift
The CEO pushes back on the idea that AI-assisted coding will cause enterprises to simply rebuild their own versions of platforms like Salesforce. His view is that major SaaS categories persist but their business models shift — moving away from per-seat pricing toward outcome-based models — while AI lowers the barrier to building ancillary internal tooling. The net result is more total software in use, not the elimination of incumbent platforms.
In fintech specifically, the multiproduct race is already well underway. The CEO notes that launching Mercury in 2019 with a single bank account product was sufficient then, but competing against Mercury today requires a bundled offering including credit cards, bill pay, and invoicing. AI accelerates product velocity but does not eliminate the craft required to build high-quality features.