I. The Second Wave: Different Markets, Different Infrastructure Requirements
The verticals where second-generation neobanks are finding opportunity — SME banking, multi-currency accounts for internationally mobile professionals, embedded finance for non-bank platforms, and regulated digital banking in GCC and African markets — share a set of characteristics that distinguish them from the consumer current account market that defined the first wave.
Regulatory complexity: SME banking involves credit underwriting, CASS client money obligations, and complex KYB requirements. GCC digital banking requires Arabic language, Sharia compliance monitoring, CBUAE or SAMA licensing, and local sanctions list integration. Multi-currency products require FX margin management, hedging, and treasury P&L reporting.
Multi-jurisdiction compliance: second-wave neobanks often seek to serve customers or operate across multiple regulatory jurisdictions from early in their development. The infrastructure must support FCA, CBUAE, MAS, and RBI requirements simultaneously — not sequentially.
B2B distribution: many second-wave neobanks distribute through API partnerships, white-label arrangements, or embedded finance integrations. This requires a BaaS-grade API infrastructure that can support third-party developers, not just a consumer mobile app.
II. The Build vs Platform Decision
The most consequential infrastructure decision a neobank makes is not which cloud provider to use or which card scheme to partner with — it is whether to build core banking infrastructure from scratch or deploy on an established platform. The strategic case for building is well-understood: control, flexibility, and differentiation. The operational reality is more challenging.
Building a production-grade, compliant multi-currency general ledger from scratch typically requires 18 to 24 months of engineering effort from a team of 8 to 12 senior engineers, with significant subsequent investment in maintenance, regulatory updates, and incident management. Based on neobank infrastructure build timelines documented in public regulatory filings and published case studies, 2022–2025²⁰
Adding IBAN issuance, card issuance, FX engine, interest calculation, open banking API, and multi-jurisdiction regulatory reporting to a self-built ledger extends the timeline by a further 12 to 18 months and materially increases team size requirements. The cumulative cost of infrastructure that a platform provides pre-built — and which provides no customer-facing differentiation — is typically in the range of £8 million to £20 million over the first three years, depending on geography and regulatory scope.²⁰
A neobank that deploys on an established platform in Q1 2026 will have 18 months of live customer data, revenue traction, and product iteration data by the time a team building from scratch completes infrastructure testing. In a market where product-market fit is discovered through iteration rather than planning, this difference is often decisive.
III. The Core Infrastructure Checklist
Multi-Currency General Ledger
The ledger is architecturally foundational — every other system depends on its accuracy and performance. For UK-regulated entities, it will typically need to support real-time position management across multiple currencies, client money segregation consistent with FCA CASS requirements, intraday liquidity monitoring, and an audit trail suitable for both internal controls and regulatory inspection. Off-the-shelf ledger solutions not built for regulated entities create compliance remediation costs that frequently exceed the initial saving.
Virtual IBAN and Account Issuance
Business customer and multi-currency neobanks require virtual IBAN issuance infrastructure. Customers expect dedicated account numbers; enterprise clients require them for counterparty payment receipt. The IBAN issuance layer must integrate with the ledger and with scheme settlement to ensure accurate posting of incoming payments — a requirement that creates significant data architecture complexity if not designed in from the outset.
FX Engine with Configurable Margin Architecture
FX revenue is typically 20 to 40% of total revenue for multi-currency neobanks — making the FX engine arguably the highest-value infrastructure component after the ledger itself. The engine must support real-time rate feeds from multiple providers, configurable margin structures by corridor and customer segment, hedging integration for FX exposure management, and treasury P&L reporting. Neobanks that launch with static, uniform FX margins leave material revenue on the table from day one.
Card Issuance Programme
Consumer and SME neobanks require card programmes for everyday spend. This involves physical and virtual card issuance with a scheme sponsor, 3DS2 authentication for PSD2 SCA compliance, Apple Pay and Google Pay tokenisation, and granular spending controls for corporate programmes. The card programme must integrate with the ledger for real-time balance deductions — a requirement that rules out disconnected card-processing arrangements common in early-stage neobank builds.
Open Banking API Infrastructure
EU and UK neobanks are generally expected to maintain production-quality open banking APIs for account information and payment initiation services, with specific obligations varying by licence type and jurisdiction. Under PSD3, API quality standards are being materially raised — including mandatory uptime SLAs and response time requirements. A neobank that launches with a minimal or unstable open banking API faces both regulatory scrutiny and commercial disadvantage as aggregators and enterprise clients increasingly weight API quality in platform selection.
IV. The AI Layer and Unit Economics
Beyond core infrastructure, the neobanks with the best unit economics in 2026 are those that have deployed AI across their commercial operations from launch. AI-driven customer health scoring identifies at-risk customers before churn, enabling retention interventions that are materially cheaper than acquisition. AI cross-sell engines surface the right product offer at the optimal moment for each individual customer — improving ARPU without increasing marketing spend. AI-personalised onboarding and engagement flows drive daily active usage in ways that generic product configurations cannot replicate.
The commercial case for embedding AI from launch rather than adding it later is compelling. The incremental cost of deploying AI capabilities on top of existing platform infrastructure is significantly lower than retrofitting AI models onto a legacy system. The data advantage — earlier accumulation of the transaction and behavioural data that trains AI models — compounds over time in favour of firms that start earlier.
360 Fintech AI provides the complete infrastructure stack for a 2026 neobank launch: multi-currency ledger, virtual IBAN issuance, FX engine, card issuance, open banking API, multi-jurisdiction regulatory reporting, and AI-powered commercial intelligence — in a white-labelled, multi-tenant platform deployable within weeks. Contact launch@360fintech.ai.