Phoenix – Reliability and Digital Experience Transformation for BirBank (KapitalBank)

The goal of Project Phoenix is to strengthen the reliability, continuity, and customer experience of all BirBank (KapitalBank) systems by directly connecting technology and business performance indicators.

Phoenix contributes to the strategic initiative “BirBank Stability Improvement & Tech Transformation”, targeting by Q4 2026:

• System Availability (Unplanned Downtime): reach 99.94 % through HA architecture, proactive monitoring, and automation.

• Process Availability (Excluding Planned Downtime): reach 99.49 % via end-to-end observability, DR readiness, and digital twins of business processes.

• Apdex Rating (Digital Experience): exceed 85 %, linking uptime and latency to customer satisfaction across all channels.

Phoenix ensures that every improvement in infrastructure and process stability directly enhances customer trust and experience.

Between 2024 and 2025, Phoenix achieved major milestones in reliability and experience metrics:

• System uptime: 99.83 % (2023) → 99.98 % (2025).

• HA coverage: 100 % (2025) for all critical systems.

• DR coverage: 90 % (2025) (target 100 % by 2026).

• Monitoring coverage: 100 % (2025) achieved across all environments.

• Apdex (Digital Experience): 0.85 (↑ trend, target > 0.90 by 2026).

• Capacity margin: 2–4× increase across all critical systems.

• QA: 24.5 K test cases, 700–800 critical bugs detected monthly before release (2025).

• Annual cost savings: 1.6 M AZN (2025) from monitoring optimization.

Phoenix turned uptime, process availability, and digital experience into connected, measurable outcomes — redefining reliability for Azerbaijan’s largest digital bank.

• Infrastructure: VMware, Kubernetes, PostgreSQL, Oracle, Redis, Pasha Tech Cloud, SD-WAN.

• Monitoring & Observability: Dynatrace (AI root-cause engine), In-house Apdex Framework, Prometheus, Grafana, Tableau.

• QA & Testing: Jenkins, Postman, Xray, Selenium, JMeter (Load Testing).

• Automation & DevOps: GitLab CI/CD, Ansible, Docker.

• Network & DR: Fortinet Firewalls, NetApp Storage, Cisco DNA, Active-Active DC Architecture (Main DC Baku + DR DC Goychay +200 km).

Phoenix faced an extraordinary level of complexity:

1. Legacy landscape: the program started from near-zero observability. A single mobile outage previously lasted up to 12 hours — unacceptable for a digital bank.

2. 20-year process gap: BirBank had no structured QA, Load Testing, or Performance Management practices — these functions had never existed in modern form.

3. Resource gap: there were no QA or LT engineers initially. The team had to build internal expertise and create a technical community from scratch.

To close these gaps, we established engineering communities and guilds for QA, Business Analysts, Developers, SRE, and DevOps, along with a dedicated AI-in-SDLC & ITOps Guild for applying AI agents to automation and root-cause detection.

Through programs like QA Tech Mastery, we trained 40+ engineers — building a new generation who see quality, reliability, and AI-driven engineering as shared responsibilities.

Phoenix is a large-scale reliability and performance transformation program covering all systems of BirBank (Kapital Bank) — Azerbaijan’s leading digital banking platform serving over six million users.

Launched in April 2024, Phoenix rebuilt BirBank’s reliability foundation and connected technology performance directly to customer experience.

The program united five core streams that together reshaped both architecture and culture:

1. Monitoring and Observability

• Built centralized observability using Dynatrace, later complemented by a fully in-house platform.

• Introduced Apdex-based digital experience monitoring, analyzing each transaction phase.

• Integrated Dynatrace AI engine (Davis AI) for proactive incident correlation and root-cause prediction.

2. High Availability and Disaster Recovery (HA & DR)

• Migrated all critical systems to HA architecture.

• Implemented active-active DR between data centers in Baku and Goychay (+200 km).

• Conducted a 24-hour DR switchover, the largest in Azerbaijan’s banking sector.

3. Quality Assurance and Load Testing

• Built QA capability from zero, integrated GUI/API testing into CI/CD pipelines, and established dedicated Load Testing environments.

4. Capacity and Performance Management

• Introduced performance baselining and continuous load validation for all high-traffic systems.

5. Engineering Culture and Resilience

• Launched chaos testing and digital twins for predictive reliability.

• Embedded reliability KPIs (Apdex, MTTD, MTTR) into squad OKRs and created communities of practice.

Phoenix transformed reliability from a maintenance function into a strategic, AI-enhanced engineering discipline, ensuring every system change can be measured through customer impact.

Phoenix covers all systems and infrastructure of BirBank (Kapital Bank) across Azerbaijan.

Two geographically distributed data centers — Main DC (Baku) and DR DC (Goychay, +200 km) — operate in active-active mode.

BirBank serves over six million retail and corporate customers nationwide.

The architecture, digital-twin model, Apdex-based monitoring, and AI-driven observability developed within Phoenix are now replicated across Kapital Bank as a national benchmark for resilience and experience management.