Automation & Integration in Stockholm

Financial Services / Fintech and Payments

A Stockholm-based fintech payment company (processing SEK 85B annually, 15,000 merchant customers, 180 employees — providing payment solutions for Nordic e-commerce and omnichannel retail) needed to automate its operations to handle 10x transaction volume growth without proportional staff increase, while maintaining compliance across four Nordic markets. (1) Merchant onboarding was slow: onboarding a new merchant took an average of 12 business days — involving KYC/AML checks, risk assessment, contract generation, technical integration setup, and payment method activation. The onboarding team (12 FTEs) was the primary bottleneck for growth — the company received 400 merchant applications per month but could only onboard 280, creating a backlog that grew monthly. (2) Transaction monitoring was overwhelming: the company processed 2.8 million transactions per day. Compliance regulations (Swedish Money Laundering Act, and equivalent regulations in Norway, Denmark, and Finland) required transaction monitoring for suspicious activity. The rule-based monitoring system generated 850 alerts per day — but the compliance team (8 FTEs) could only investigate 120 alerts daily, meaning 86% of alerts were acknowledged but not investigated. Regulators had flagged this investigation gap in the last Finansinspektionen examination. (3) Settlement and reconciliation: merchant settlements required matching transactions with bank clearing data across 4 Nordic payment networks (Bankgirot in Sweden, BankAxept in Norway, Nets in Denmark, and Finnish payment clearing). Each country had different settlement cycles, formats, and exception handling requirements. Settlement reconciliation consumed 6 FTEs — with discrepancies averaging 0.8% of transaction volume requiring manual investigation. (4) Multi-country tax compliance: the company operated as a payment facilitator — requiring VAT reporting in all 4 Nordic countries. Swedish Skatteverket, Norwegian Skatteetaten, Danish Skat, and Finnish Vero each had different reporting formats, deadlines, and requirements. The tax team (3 FTEs) spent 80% of their time on compliance preparation rather than tax planning. (5) Merchant support was reactive: merchants contacted support for transaction disputes, settlement enquiries, and technical issues — 2,800 support tickets per month. Support staff (15 FTEs) spent 40% of their time on enquiries that could be answered from available data (settlement status, transaction details, refund status) but required manual lookup across multiple systems.

We built a comprehensive operations automation platform over 12 weeks. (1) Automated merchant onboarding: digital application — merchants submitted structured applications through a self-service portal. BankID verification: identity verification of beneficial owners through BankID (Sweden), BankID (Norway), NemID/MitID (Denmark), and Finnish Trust Network. Automated KYC: company registry checks (Bolagsverket in Sweden, Bronnoysundregistrene in Norway, CVR in Denmark, PRH in Finland), sanctions screening, PEP checks, and adverse media monitoring — all automated with results aggregated into a risk score. Risk decisioning: low-risk merchants (established businesses, standard risk profile) approved automatically. Medium-risk merchants flagged for expedited manual review with pre-populated risk assessment. High-risk merchants routed to the compliance team with detailed investigation brief. Technical setup: automated payment method activation — API credentials generated, test environment provisioned, and integration verified through automated test transactions. (2) AI-powered transaction monitoring: ML models replacing rule-based monitoring — trained on 3 years of transaction data including confirmed SAR cases. Behavioural analysis: monitoring transaction patterns against merchant profiles — detecting anomalies in volume, velocity, amount distribution, and geographic patterns. Network analysis: identifying connected merchants and transaction patterns indicating money laundering, fraud, or sanctions evasion. Alert prioritisation: ML-scored alerts ranked by risk severity and confidence — compliance team investigated the highest-risk alerts first. Alert volume: ML models reduced actionable alerts from 850 to 180 per day (79% reduction) while detecting 40% more genuinely suspicious activity. (3) Nordic settlement automation: unified reconciliation engine matching transactions against clearing data from all 4 Nordic payment networks. Country-specific adapters: handling Bankgirot, BankAxept, Nets, and Finnish clearing formats through normalised processing. Automated matching: 99.6% of transactions matched automatically (from 99.2% — the improvement may seem small, but at 2.8M transactions daily, the 0.4% improvement eliminated 11,200 daily manual reconciliation items). Exception handling: unmatched transactions automatically classified by exception type (timing, amount, reference) with recommended resolution — reducing manual investigation time by 70% for remaining exceptions. (4) Multi-country tax automation: automated VAT calculation and reporting for all 4 Nordic markets — country-specific rules applied based on merchant location, transaction type, and destination. Skatteverket: automated Swedish VAT return preparation (Skattedeklaration) and SAF-T export. Pan-Nordic: equivalent automation for Norwegian MVA-melding, Danish momsangivelse, and Finnish ALV-ilmoitus — each generated from the same transaction data but formatted per country-specific requirements. Deadline management: automated calendar with preparation workflows triggered ahead of each country's filing deadline. (5) Merchant self-service: automated merchant portal — real-time access to transaction data, settlement status, refund processing, and dispute management. Self-service resolution: merchants resolved 65% of their enquiries through the portal without contacting support — settlement details, transaction search, and refund initiation available 24/7. Automated notifications: proactive alerts for settlement completion, dispute deadlines, and integration issues — reducing the need for merchants to contact support for status updates.

Merchant onboarding: from 12 business days to 2 business days (83% reduction). Low-risk merchants onboarded in under 4 hours (fully automated). Onboarding capacity: from 280 to 600+ merchants per month with the same team — clearing the backlog and enabling growth without hiring. Compliance: Finansinspektionen examination finding resolved — all alerts now investigated or triaged within 24 hours. SAR filing rate increased by 40% (more suspicious activity identified through ML). False positive rate reduced by 79% — compliance team focused on genuine risks rather than noise. Settlement: automated matching rate from 99.2% to 99.6% — eliminating 11,200 daily manual items. Settlement team from 6 FTEs to 2 FTEs — remaining team handled complex cross-border exceptions. Settlement error rate: zero merchant-impacting errors in 12 months (from 45 per year previously). Tax compliance: multi-country VAT reporting fully automated — tax team from 3 FTEs to 1 FTE focused on planning rather than compliance preparation. Filing accuracy: zero corrections required across 48 quarterly filings (4 countries x 4 quarters) in the year following implementation. Merchant support: self-service portal resolved 65% of enquiries without human contact. Support ticket volume from 2,800 to 980 per month (65% reduction). Support team from 15 FTEs to 8 FTEs — remaining team handled complex technical and dispute cases. Annual value: SEK 145M (growth enablement — additional merchant onboarding driving SEK 28B additional transaction volume at average 0.52% margin), SEK 52M (staffing efficiency — 26 FTEs redeployed across onboarding, settlement, tax, and support teams at average SEK 750K fully loaded cost), SEK 28M (compliance risk — resolved regulatory finding, avoided potential licence conditions, and reduced fraud/AML exposure), SEK 18M (merchant retention — improved experience driving 4% retention improvement on existing merchant base), SEK 8M (settlement accuracy — eliminated errors, reduced bank charges, and improved cash flow predictability) = SEK 251M (approximately USD 24M). Development cost: SEK 3.2M (approximately USD 310K). Payback: under 5 days.