[ AI agency · sector banking ]

AI agency
for banking, fintech and wealth management.

Faster KYC. Fraud detected sooner. Reporting without Saturdays.

We apply AI in Spanish banking, fintech and wealth management for KYC/onboarding, fraud detection, regulatory reporting and portfolio analysis. With architecture ready for DORA and MiFID II, in European cloud or on-premise.

80 %
onboardings automated end-to-end
×5
fraud detected vs baseline
−60 %
regulatory reporting time
[ Sound familiar ]

The 4 problems we see in every banking, fintech and wealth management in Spain.

/ 01

Manual KYC/onboarding slowing acquisition

Client abandons if it takes >48 h to operate. Manual verification, AML review and signature scattered across 5 people.

/ 02

Fraud detected late and expensive

Fixed-rule systems that don't catch new patterns. Each undetected fraud case is direct loss and potential fines.

/ 03

Regulatory reporting consuming whole teams

DORA, Solvency, PSD2 SCA, MiFID II. Reports built by hand with Excel and cross-validation. Doesn't scale.

/ 04

Portfolio analysis and advisory not personalised

Standard recommendations by profile, not by specific client. EAFs and private bankers want to personalise but lack time.

[ How we apply AI in your sector ]

Real use cases for banking, fintech and wealth management.

Not a list of technologies. Concrete processes where AI delivers measurable euros or hours from the first month.

01

Automated KYC/AML with AI

Identity verification, AML, sanctions lists, client risk analysis. Escalation to compliance only for high-risk cases.

02

Fraud detection with adaptive models

Real-time transaction analysis. Detects new patterns fixed rules miss. Feedback loop with fraud team.

03

Automated DORA / MiFID II reporting

Consolidation from multiple systems, report generation in regulator format, anomaly alerts before sending.

04

AI portfolio analysis

Personalised recommendations per specific client, rebalancing alerts, exposure analysis and event sensitivity.

05

Copilot for bankers and EAFs

Assistant with access to portfolio, contact history, commercial proposals. Drafts emails, prepares meetings, summarises sector news.

06

Voicebot and chatbot for first-level

Balance queries, movements, cards. Enhanced PSD2 authentication before sensitive operations.

[ Recommended services ]

Our services ranked by impact for banking.

[ Verisimilar cases ]

Figures, sector, result.

Invented cases with metrics consistent with our real ranges, until a client authorises publishing their own.

01
Spanish neobank · 250k clients

Automated KYC with AI

Identity verification + AML + sanctions lists in under 3 min. Escalation to compliance only for high-risk cases.

−72 % onboarding time
02
EAF · Madrid · €80M AuM

Copilot for advisor with portfolio access

Assistant with access to portfolios, history and market. Drafts personalised emails, prepares meetings, alerts rebalancing.

8 h/week saved per advisor
03
Consumer lending fintech · 15k operations/month

Real-time fraud detection

Adaptive model trained on 3 years of history. Detects new synthetic identity patterns rules don't catch.

×5 cases detected vs fixed rules
[ Regulation and compliance ]

We work with sector regulation, not against it.

Our pilots incorporate from design the regulatory requirements that apply to the banking sector in Spain:

[ FAQ ]

What the sector asks.

How do you comply with DORA on an AI project for banking?

DORA (EU Regulation 2022/2554) requires ICT risk management, incident management, resilience testing, third-party management and information sharing. Any AI component we deploy enters the bank's ICT third-party register.

We design with criticality classification, fallback plans if the model fails, human escalation procedure, and audit documentation ready for supervisor inspection (BdE, CNMV, ESMA, EBA).

For critical components (KYC, fraud, credit decisions) we deploy in your infrastructure or European cloud with Spain residency. Never public API for regulated data.

Does a chatbot comply with PSD2 SCA for sensitive operations?

The chatbot itself is not the element complying with PSD2 SCA (Strong Customer Authentication). The chatbot converses; for any payment operation or sensitive data modification, your existing SCA flow triggers (biometrics, OTP, possession factor).

We design the flow so the transition conversation → authentication → operation is smooth (e.g., bot opens app to complete SCA). This complies with PSD2 and gives modern client experience.

Purely informational queries (balance, last transactions, card blocked yes/no) require standard channel authentication, not SCA.

Can you work with LLM on-premise for data that can't leave the bank?

Yes. For clients with strict data residency requirements we deploy open-source models (Llama, Mistral) in your on-premise infrastructure or in European private cloud. Data never leaves the bank's perimeter.

Current open-source models cover most enterprise use cases with competitive performance. When the use case requires commercial models (Claude Opus, GPT-5), we use enterprise options with no-retention data contracts and European cloud.

Open source vs closed model decision is made with you case by case, balancing quality, cost, latency and regulatory requirements.

How long for a pilot and what ROI can you expect?

Automated KYC: 8-12 weeks, €30-70k. Typical ROI: CAC reduction (less abandonment from slow onboarding) + compliance operational cost.

Fraud detection: 10-16 weeks including history training, €40-90k. ROI: every 10 bp of extra fraud detected over baseline usually covers annual cost with margin.

Automated regulatory reporting: 6-10 weeks per report, €15-30k per report. ROI: freeing compliance FTEs for value tasks.

[ Keep reading ]

Start with an acotated pilot?

Free 30-min diagnosis. We come out with 3-5 use cases prioritised by ROI for your specific financial entity.