As Bangladesh’s capital market pushes toward full digitization, the technology behind the trading screen has never mattered more. Munshi Mustafijur Rahman, deputy general manager and head of IT infrastructure at Dhaka Stock Exchange (DSE), speaks to Dhaka Tribune’s Wafiur Rahman about building a future-ready exchange.
You have spent 12 years at DSE. How would you summarize the journey?
I joined in 2014 in the System and Market Administration department, looking after the core trading system, and over time moved into infrastructure leadership.
The journey has been about steadily modernizing the exchange’s technology backbone while keeping the market running without interruption.
The milestones I am proudest of: the Rated-3 data centre at DSE Tower and the near-disaster-recovery site, moving our core Order Management System to a new virtualization platform that became the exchange’s largest annual cost saving, developing the Virtual NOC monitoring platform, and implementing an OpenStack-based financial private cloud (IaaS) with our partner exchange.
Overall, it has been a transition from a traditional, hardware-dependent setup toward a resilient, automated and digitally mature infrastructure.
You are known as an open-source advocate. What drew you to it in a cautious, financial-market environment?
Financial markets are cautious for good reason: reliability and security come first. But open-source technology today is enterprise-grade; much of the world’s critical infrastructure runs on it.
What attracted me is the combination of flexibility, transparency and cost efficiency. It lets us avoid licensing lock-in, understand exactly what is running in our environment, and automate at scale.
The key is discipline: open source still needs professional support, skilled people, testing and governance.
Used that way, it gives a financial institution both innovation and control. As the first RedHat Certified Instructor (RHCI) and RedHat Certified Architect (RHCA) in Bangladesh, I have seen that discipline work.
How should a legacy exchange approach modernization?
As a phased transformation, never by replacing everything at once. The goal is a highly available, scalable, automated and fully digitized market infrastructure, achieved without disrupting trading.
Technology has evolved from physical servers to virtualization, then cloud, then containers and automation.
Today, microservices and containerization, which means packaging an application so it runs consistently on any platform, are central to fintech infrastructure because they let you introduce change with minimal impact on what is already running.
In practice, I would start with the core Matching Engine, gradually transforming monolithic systems into independently manageable services.
Then the Order Management System, progressively re-engineered on containerized technologies, with locally developed fintech solutions considered where they meet our performance and security requirements. Then the investor-facing digital services.
With so much change, how do you keep the platform reliable?
Reliability has to be engineered in; you cannot achieve zero downtime simply by adopting new technology.
It comes from redundancy at every layer, clustering with automatic failover, and disaster-recovery capability across sites, so no single failure can interrupt trading.
Automation and real-time observability are essential, and Infrastructure as Code, meaning environments defined and managed through code rather than manual setup, makes everything standardized and repeatable.
Moving the core OMS from VMware to DCS virtualization reportedly delivered major savings. How?
DCS is our new enterprise-grade virtualization platform, and I was involved from planning through production deployment.
Since the OMS is mission-critical, the priority was transition without disrupting trading.
Moving off VMware and optimizing our licensing model produced significant annual savings, and DCS gives us a more scalable platform with greater automation in daily operations, reducing manual effort and configuration errors.
That sounds high-stakes. Was there a moment that tested you?
The final cutover weekend was the most intense period of my career.
When you migrate the system that runs the country’s stock trading, the market must open on time the next trading day, full stop.
We rehearsed for weeks in a test/UAT environment, documented every step, and kept a complete rollback plan.
The team stayed through the night validating each stage. When the market opened normally the next morning, with traders unaware anything had changed, it was the best compliment an infrastructure team can receive.
The lesson: in mission-critical technology, success is invisible; you earn it through preparation, rehearsal and having a way back at every step.
DSE’s technology has faced public criticism, including trading disruptions. What do you say to investors?
Criticism comes with the territory, and investors have every right to demand reliability.
What I can say is that the exchange has invested in exactly the areas that determine it: a certified Rated-3 data centre, dedicated disaster recovery, modernized virtualization, centralized monitoring and hardened systems.
No exchange anywhere can promise nothing will ever go wrong; a serious exchange promises that failures are anticipated, contained and recovered quickly, and that every incident makes the system stronger.
Resilience is now designed into our architecture rather than added afterwards.
You built a custom monitoring platform, the Virtual NOC. What is it?
Banks and telecoms run 24/7 physical Network Operations Centres.
I developed the Virtual NOC, or vNOC, registered with the Bangladesh Copyright Office, as a cost-effective way to do the same job: one centralized, real-time view of servers, network devices, storage, firewalls, load balancers, UPS systems and other data centre components, raising automatic alerts (Voice call, SMS and email) the moment anything behaves abnormally.
Alongside it, server hardening keeps the systems themselves secure: tightly configured, patched, access-controlled.
Monitoring and hardening together give us early detection and prevention.
Looking ahead, how will AI, high-performance computing and blockchain shape the market?
AI-driven surveillance will make market monitoring far more intelligent, analyzing large volumes of trading data and flagging potential abuse faster than predefined rules ever could, and identifying suspicious or potentially fraudulent activities before they occur.
High-performance computing will matter as data volumes grow and markets demand very low latency.
Blockchain’s real potential here is in post-trade settlement, asset ownership and reconciliation, but it should be adopted for specific problems where it clearly outperforms a conventional database, not because it is fashionable.
And for Bangladesh, data sovereignty, cybersecurity and regulatory compliance must anchor all of it.
What infrastructure does a next-generation exchange need?
A certified data centre of at least Tier III or Rated-3 standard, with redundancy across power, cooling and connectivity.
Then two layers of disaster recovery: a near-DR site close to the primary for fast, near-synchronous recovery, and a far-DR site at geographic distance (different tectonic plate) to survive a regional disaster such as an earthquake, fire, flood or grid failure that could affect both.
Add software-defined, scalable computing, AI-enabled cybersecurity and strong governance over where sensitive market data is stored and processed.
A future-ready exchange is not a bigger data centre; it is a resilient ecosystem in which all of these work together.
Finally, how do you see DSE, and yourself, in 10 years?
I would like to see DSE among the region’s most technologically advanced exchanges: fully digital, cloud-native, with AI-driven surveillance and cyber-resilience built into everything, enabling new products, instant settlement and greater transparency for investors.
For myself, I see the role evolving from building infrastructure to shaping strategy, mentoring the next generation of engineers, and building local capability.
If, 10 years from now, DSE runs on a resilient, largely home-grown platform and the team I helped build is leading it confidently, I will consider that a success.