When HBL began migrating onto the Temenos⁠ core banking platform, the bank was not attempting a conventional replacement exercise. The programme involved modernising operations supporting the bank’s branch network, over 40 million customer accounts and approximately 20 million transactions daily across conventional and Islamic banking environments.

The first phase of deployment covered around 200 branches, with the broader programme progressively expanding across Pakistan. The implementation also involved hybrid-cloud deployment powered by Red Hat OpenShift⁠ and delivery support from Systems Limited⁠.

For Faisal Anwar, HBL’s chief technology officer (CTO), the transformation was driven less by front-end digitisation than by the operational burden of managing increasingly complex banking environments at scale.

“The primary focus of this transformation revolves around a couple of things,” Anwar said. “One is what we drive out as an empathy for our employees.”

He argued that simplifying core systems reduces operational friction across frontline and back-office environments while creating the foundation for progressively more autonomous operating models over time.

“We will adopt the inclusion of AI Ops into it,” Anwar said. “The core becomes a lot self-healing and self-monitoring, and humans are there then to supervise.”

The broader objective was therefore not simply replacing infrastructure. HBL was attempting to build a banking environment capable of remaining continuously operational while becoming more modular, cloud-ready and artificial intelligence (AI)-enabled simultaneously.

“We wanted to create agility and digitally native bank,” Anwar said.

Running coexistence without disrupting customers

A central operational challenge throughout the programme has been coexistence between legacy and newly modernised environments.

Rather than attempting a large-scale migration across all branches simultaneously, HBL adopted a phased conversion strategy intended to minimise operational disruption and narrow migration risk exposure.

“If I were to cover the bank’s branch network in a single core as a big bang, that means I need to take an outage of couple of days,” Anwar said.

Instead, the bank progressively converts selected branch groups while maintaining servicing continuity across neighbouring locations and broader operational networks. The sequencing itself became operationally strategic.

“For example, two neighbouring branches, we ideally would not want to convert at the same time,” Anwar said. “If something happens, we can still serve a customer from a neighbouring branch.”

The approach substantially reduces the operational blast radius of each migration cycle while allowing the bank to continue normal operations during transformation.

At the same time, coexistence introduces significant architectural and operational complexity. “It’s reasonably complex,” Anwar said. “What you’re trying to replicate is a seamless frictionless experience to the customers.”

HBL operates across multiple customer channels, including mobile banking, ATMs, branch systems and agency banking networks, all of which must continue operating consistently while parts of the underlying core environment are modernised.

Anwar said much of the operational effort involved scenario planning, dependency mapping and extensive operational whiteboarding to anticipate edge-case failures before migration cycles occurred.

“There was a lot of kind of whiteboarding,” he said. “Writing down the possible scenarios and then working around them.”

Decomposing the bank into operational domains

One of the strongest architectural themes throughout the discussion was decomposition.

Rather than maintaining heavily monolithic operating structures, HBL has progressively decomposed parts of the bank into modular operational domains connected through middleware, application programming interfaces (APIs) and service frameworks.

“What we have done is we now have an architectural blueprint which has decomposed the bank into multiple spaces,” Anwar said.

The objective is operational isolation rather than complete dependency across the stack. “If I want to work on my front end, I can work without having to impact my back end,” he said.

The same logic increasingly applies to payments infrastructure. HBL is evaluating whether payment systems should remain embedded within the core environment or operate independently as decomposed modules capable of being upgraded separately.

“Keep the core lean,” Anwar said.

Under that approach, operational disruption becomes more containable. If one component becomes unavailable, the remainder of the bank continues functioning rather than cascading into wider service interruption.

“So if the credit card system is down, don’t worry about it,” Anwar said. “The app should still work.”

The architecture separates operational continuity from complete system dependency.

“It is not a large-scale disruption because it’s a decomposed architecture,” he said.

Transformation expands the security surface

Anwar argued repeatedly that transformation itself creates new operational vulnerabilities.

“When you enter into transformation, there is a large shift,” he said. “The surface of attack has changed, and the vector of attack has changed.”

As banks modernise, integrate APIs, introduce cloud-native architectures and expand external connectivity, the attack surface expands materially compared to traditional monolithic environments.

“You become seriously vulnerable in a very short period of time,” Anwar said.

HBL increasingly views operational resilience, cybersecurity posture and AI-enabled defence systems as inseparable components of transformation itself.

“We need tools which are better intelligent,” he said. “Self autonomous.”

Anwar argued that AI is now materially changing the scale and speed of cyber threats confronting financial institutions.

“Earlier in the past, it was like a man who was trying to hack a machine,” he said. “Right now, it’s a machine working against a machine.”

The bank’s discussions with infrastructure and implementation partners focus increasingly on building resilient operating environments capable of adapting continuously as transformation progresses.

Operational modernisation and security architecture are therefore converging into the same strategic problem.

Keeping the core lean while personalising banking

Anwar argued that banks are becoming progressively less tolerant of excessive core customisation despite growing pressure for personalised customer experiences.

“The more you customise, the more complex you create the scenario,” he said.

HBL increasingly prefers maintaining standardised core environments where possible while shifting personalisation logic into surrounding operational and data layers. “You would rather leverage your auxiliary infrastructures to hyper personalise versus go into your main core and try to hyper personalise it,” Anwar said.

The architecture allows customer experiences to evolve independently from the core platform itself, reducing the operational burden associated with deeply customised legacy environments.

At the same time, Anwar argued that banks cannot operate purely as standardised utilities detached from changing customer behaviour. “It’s no more about conventional banking,” he said. “It’s about lifestyle banking.”

The challenge increasingly becomes balancing standardisation, upgrade discipline and operational simplicity against customer expectations for real-time, personalised and continuously available financial services.

“The more you customise, the complex it gets,” Anwar said. “I would rather keep myself updated with cutting-edge technology.”

Building infrastructure for continuous availability

Hybrid-cloud deployment has become another foundational layer within HBL’s broader operating model.

The bank deployed its environment on Red Hat OpenShift specifically to create cloud-native and cloud-ready infrastructure capable of supporting future scalability and resilience requirements.

“The way we are setting up the bank infrastructure is that it’s going to be cloud native, cloud ready,” Anwar said.

The architecture allows operational flexibility without requiring wholesale redesign of applications or infrastructure environments.

“If I want to lift, it’s just a lift and shift,” he said.

HBL is also evaluating multi-site operating environments to improve concurrent maintainability and operational continuity.

“If I want to upgrade my one site, I should not be taking a downtime,” Anwar said.

The objective increasingly becomes continuous operational availability even while infrastructure upgrades, maintenance and transformation activities continue underneath the operating environment itself.

Banks are therefore moving away from infrastructure environments dependent on maintenance windows and planned operational interruption.

AI readiness is becoming an infrastructure challenge

Although AI dominated discussions throughout Temenos Community Forum 2026, Anwar argued that infrastructure readiness matters more than experimentation alone.

“I would say we are at a moderate level and moving towards maturity,” he said.

HBL has already deployed AI-ready infrastructure components and data environments, including the Temenos Data Hub acting as a centralised repository for downstream data consumption.

The bank focuses less on isolated AI use cases and more on operational AI domains such as AI Ops, AI security and engineering productivity. “AI Ops to simplify our operations,” Anwar said.

HBL is also evaluating how AI can accelerate engineering workflows across business requirements, functional specifications and code generation.

“What I see is that core engineering over a period of time will now translate into prompt engineering,” Anwar said.

At the same time, the bank remains cautious around uncontrolled deployment of open-source AI environments and unsecured code generation.

“What we don’t want is a code generated which is vulnerable enough to create another problem for us in the bank,” he said.

For HBL, AI readiness increasingly depends less on models themselves and more on whether banks possess sufficiently modern infrastructure, operational governance, security architecture and data environments to absorb AI safely at scale.

Continuous transformation is becoming the banking model

Throughout the discussion, Anwar repeatedly returned to the idea that banking transformation itself is changing structurally.

Historically, many large banks approached modernisation as periodic programmes separated by long periods of relative stability. HBL increasingly operates on the assumption that transformation itself will remain continuous.

That changes the role of architecture, operational resilience, upgrade discipline and technology partnerships simultaneously. Systems increasingly need to remain modular, cloud-ready, continuously operable and upgradeable while supporting real-time customer expectations and increasingly AI-enabled operating environments.

HBL’s relationship with Temenos and Systems Limited therefore functions not simply as a vendor engagement but as a long-term operational partnership around continuous technology evolution.

“As a CTO, you want to have your good night’s sleep,” Anwar said. “That good night’s sleep would only come with trust.”

For large incumbent banks, the challenge is no longer whether transformation should happen. Increasingly, the challenge is whether institutions can continue modernising continuously without destabilising the systems already supporting customers, payments, compliance and operational resilience every day.