About Me
In 2006, at age 14, I was featured on the front page of Prothom Alo — Bangladesh's largest newspaper — as the youngest known web developer in the country. I've been building for the web ever since.
I'm Asaduzzaman "Asad" Pavel. I've spent the last decade making sure that 'millions of users' is a metric, not a crisis. I specialize in building backend systems that are stable enough to be boring, though I still enjoy shipping the occasional full-stack project when I want to see how the user actually interacts with the data.
What I Do
I've spent most of my career in the backend, usually cleaning up "scalable" systems that weren't actually built to scale. I handle the infrastructure and microservices, but I keep my frontend skills sharp enough to build full-stack apps that don't look like they were made by a backend developer.
- High-Availability Systems: Go and Node.js services that handle millions of requests without waking me up at 3 AM.
- Infrastructure as Code: AWS/GCP environments with Terraform and Kubernetes. If it's not in code, it doesn't exist.
- Pragmatic Full-Stack: Fast, responsive UIs with TypeScript and Svelte/React. I know how to build what users actually click.
- Reliable APIs: REST and gRPC patterns that other developers enjoy using, not just tolerate.
Tech Stack
Tools I use to build reliable systems
My Journey
I've worked with teams in London, Singapore, New York, and Australia. Along the way, I've:
- Led engineering teams by focusing on pair programming and clear technical paths rather than just management overhead.
- Integrated over 150 crypto exchanges for a global market data platform, building the high-throughput engine needed to aggregate real-time price and volume data.
- Reduced system downtime by 60% by gutting "noise" in the monitoring. I learned that if every alert is critical, nothing is.
- Founded a bootstrapped startup that turned a $500K profit—learning more about ROI and "good enough" code there than in any corporate role.
- Shipped video on-demand streaming and voting systems for organizations like the Australian Academy of Cinema and Television Arts (AACTA), where reliability wasn't optional.
Recent Posts
Latest thoughts on engineering and building
The Only Prometheus Metrics I Actually Alert On
Essential Prometheus metrics for applications, servers, and databases. What the Golden Signals miss and why your Alertmanager config is probably wrong.
The Meta Tags That Matter (And the JSON-LD That Gets You Cited)
I thought meta tags were optional until my articles kept getting outranked by thinner content. Turns out Google was speaking a language I was not.
Go 1.26 new(expr) and the pointer problem I thought was solved
The new(expr) syntax in Go 1.26 should have let me delete my generic ptr[T] helpers. It almost did.
Beyond the Code
I'm usually the one arguing for the simpler solution. I've seen enough "elegant" systems fall apart under actual traffic to know that reliability beats cleverness every time. It's not just about the code you write; it's about building systems you don't have to fix at 3 AM.
I assumed early on that more complexity meant more value, but it's usually the opposite. I'd much rather ship a stable, "boring" system that works than a perfect one that never sees the light of day. It's the only way I've found to build things that actually scale.
If you're looking for an engineer to help build or scale your stack without turning it into a maintenance nightmare, let's talk.