What tech stack does Citigroup use?

Citigroup's core banking systems are built on decades of Java and C++ codebases, supplemented by Python for quantitative and data science workflows. The bank runs a hybrid architecture: a significant legacy mainframe footprint for core transaction processing, surrounded by increasingly cloud-native microservices and containerized applications managed via Kubernetes and Terraform. Engineering job postings on Citi Careers confirm Harness for CI/CD pipeline management, Terraform for infrastructure-as-code provisioning, and Kubernetes for container orchestration.

The defining infrastructure move of 2024–2026 is Citi's multi-year strategic agreement with Google Cloud, announced October 28, 2024. Under this deal, Citi migrates risk models, workforce applications, and customer-facing systems to GCP. The partnership includes adoption of Google Vertex AI for generative AI use cases (developer toolkits, document processing, customer service) and Google Agent Assist (voice AI for customer service agents), and leverages GCP's high-performance computing capabilities to execute millions of risk computations daily in Citi's Markets business. Citi also runs workloads on AWS and Azure as part of a multi-cloud strategy, though Google Cloud is the primary strategic partner. Akamai provides CDN and web application firewall services for Citi's public-facing web properties.

For CI/CD and infrastructure-as-code, engineering job postings confirm Terraform for provisioning and Harness for pipeline management. The Warsaw, Poland engineering hub has grown significantly as the cloud-native modernization accelerates, absorbing cloud platform, data engineering, and backend roles that support the GCP migration.

Citi's public web properties are detected using JavaScript, Bootstrap, and jQuery for frontend rendering, with Akamai providing edge delivery and web application firewall capabilities, and Ensighten for tag management and marketing analytics — all confirmed via StackShare's public Citi company profile. Citi's consumer mobile applications run on Swift for iOS. These are the publicly observable web-layer signals; the internal application stack across 200 million customer accounts is substantially more complex and not fully disclosed.

For market data, trading desks rely on Bloomberg Terminal — standard across institutional banking — along with the proprietary Citi Velocity platform for FX, rates, and securities data delivery to clients. High-performance computing clusters (migrating to GCP's HPC infrastructure) run Citi's risk models across its $2.66 trillion balance sheet. Citi Wealth has deployed Citi Sky, an AI agent built on Google's Gemini Enterprise Agent Platform, to enhance wealth advisor productivity — confirmed in the Google Cloud partnership announcement and subsequent Citi investor communications. Internal CRM systems serve the large institutional sales force, though specific vendors at this scale are not publicly disclosed.

On the AI front, Citi's O&T organization under Tim Ryan is deploying generative AI tooling broadly: Vertex AI for software development assistance (coding copilots for engineers), document processing automation for trade finance and compliance workflows, and Agent Assist for contact center automation. CEO Jane Fraser noted in 2025–2026 that AI adoption is a core pillar of the post-transformation technology strategy, with 90% of transformation programs at or near target state as of Q1 2026.

Citi's declared Cloud First strategy and its Google Cloud anchor deal create clear vendor opportunity signals. The bank is actively procuring cloud-native tooling, AI/ML platforms, developer productivity tools, and data infrastructure compatible with GCP. Vendors with native GCP integrations or Google Marketplace listings will face lower friction in Citi's procurement cycle. The $30 billion buyback announced at the May 2026 Investor Day signals capital discipline — technology that eliminates manual process or measurably improves operating leverage will clear internal business cases more readily than incremental headcount.

The multi-cloud posture (GCP primary + AWS + Azure) means Citi evaluates platform-agnostic vendors positively — particularly those that support data portability and interoperability across cloud providers. Python-native data science and ML tooling aligns well with Citi's quant and risk teams. On the displacement side, Citi is actively modernizing away from legacy mainframe dependencies — creating opportunity for core banking modernization platforms, event streaming middleware, and API management layers that can bridge legacy and cloud-native systems during the multi-year migration.

Security certifications are non-negotiable baseline requirements: SOC 2 Type II, ISO 27001, and FFIEC compliance are the minimum for financial services software. Engaging at the engineering or architecture level — through O&T's cloud platform or data teams — before formal procurement is the most effective path for new technology vendors. Expect multi-year enterprise contracts, formal RFPs, security assessments, and legal reviews coordinated through the vendor management office for any meaningful deal.