ING's Autonomous Tech Strategy for Finance and AI

A- A+

Serving over 40 million customers, ING developed an AI-focused, regulation-compliant autonomous architecture strategy powered by thousands of engineers to avoid vendor lock-in.

ING's Autonomous Tech Strategy for Finance and AI - bimakale.com

07 Eylül 2026 Pazartesi - 19:04 (6 Saat önce) 3 dk okuma

As a financial institution serving over 40 million customers worldwide, ING is reshaping its technological infrastructure without relying on a single vendor. In an interview with Marco Eijsackers at its headquarters in Amsterdam, the bank outlined an autonomous architecture approach designed to preserve long-term resilience and innovation speed.

Autonomous Architecture and ING's Approach

By converting its existing IT environment into a modular structure, the company aims to gain the capacity to seamlessly transition between different cloud providers and open-source solutions. In this way, risks such as service disruptions or price hikes from a single technology vendor can be isolated without impacting overall system integrity.

Strategy to Avoid Vendor Lock-in

ING's engineering teams adopted a multi-cloud architecture, creating a codebase capable of running the same functionality across different platforms. Such a structure allows AI models to be tested simultaneously on both public clouds and private data centers. Avoiding dependency on a single vendor not only provides cost advantages, but also enables rapid adaptation to regulatory changes.

Financial Regulations and Rapid Innovation

Strict regulations in the financial sector require high security around data retention, access control, and audit trails. To meet these requirements, ING developed policy-as-code and automated compliance checks. Automation minimizes human error by automatically updating relevant system components whenever a new regulation is introduced.

On the other hand, shortening technology cycles necessitate deploying new AI models into production multiple times a year. To manage this speed sustainably, ING integrated CI/CD (Continuous Integration / Continuous Delivery) pipelines, model validation steps, and performance monitoring tools. Consequently, when a new algorithm is produced, the compliance of not just the code, but also the data and model, can be verified instantly.

Recommendations for Other IT Leaders

During the interview, it was emphasized that ING's strategy could serve as a model for other large-scale organizations. In particular, leaders should pay attention to these three areas:

Modularity: Designing system components as independent services (microservices) reduces the risk of vendor lock-in.
Automated Compliance: Defining regulatory changes within code and validating them via automated tests accelerates the audit process.
AI-First Approach: Positioning artificial intelligence applications in an infrastructure-independent layer fosters innovation while enabling cost control.

While shaping ING's 2026 vision, these principles can also offer a roadmap for other industry players. In a high-risk domain like banking, striking a balance between technological flexibility and regulatory compliance is becoming the key driver of long-term competitive advantage.

In conclusion, ING's autonomous architecture strategy offers not merely a technical transformation, but a framework that guarantees business model sustainability. This approach creates an ecosystem where big data, AI, and strict regulatory environments coexist, ensuring readiness for the industry's future demands.

Source: Red Hat Blog

Kaynak: Red Hat Blog

Alakalı İçerikler

Risk‑bilinçli AI dağıtımında model kartlarının eksikleri 4 Gün önce
Red Hat, yapay zeka modellerinin risk‑bilinçli dağıtımı için model kartlarının yetersiz kaldığını vurguluyor; zorlayıcı senaryolara dair kanıt talep ediyor.
- tr
- ar
Security and Monitoring Integration Halts Crypto Mining 4 Saat önce
Separate data streams for operations and security teams can miss attacks such as hidden crypto mining; an AI‑powered unified platform closes this gap.
- tr
- en
Fraud Decisions Accelerated with Shared Context Layer 5 Saat önce
In fraud detection, delays in data flow render decisions ineffective. A shared context layer consolidates all information in one place, speeds up risk management, and enhances security.
- tr
- en
A New Security Approach Needed for Multi‑Model Environments 20 Saat önce
Docker Blog highlights that future systems operating with multiple models and multiple harnesses will outgrow current security mechanisms and require a new governance model.
- tr
- en
Redis Bağlam Katmanı AI Ajan Performansını Zedeliyor 1 Gün önce
Redis blogunda AI ajanlarının bağlam katmanına yeterli bütçe ayrılmadığını ve bunun iki tur önceki konuşmayı unutma, gereksiz token harcaması gibi sorunlara yol açtığını vurguluyor.
HashiCorp Vault’un AI ajanları için yeni IAM desteği yayında 1 Gün önce
HashiCorp Vault Enterprise, AI ajanları için kimlik ve erişim yönetimini güçlendiren yeni IAM özelliklerini genel kullanıma sundu; kurumların gizli verileri koruma kapasitesi artıyor.