SureshDhakal
Director of Engineering in Bangkok. I make platforms faster, cheaper to run and ready to scale, and I still ship code most weeks.
This year that has meant lower cloud costs, quicker response times, services that absorb traffic spikes, and AI agents built into how the team ships. See what I’m working on
About
I lead engineering in Bangkok and spend most of my time at the platform layer, where architecture, cloud, data and delivery meet.
I started in Kathmandu in 2012 and moved to Bangkok in 2015. Along the way my work moved down the stack, from interfaces to services to the infrastructure and data underneath. That is where most of a system’s cost, speed and reliability is decided.
Lately a large part of the job is AI-assisted engineering. That means agents wired into the delivery pipeline, not a chatbot on the side. A well-specified ticket becomes a draft pull request in minutes, and engineers validate it, review it and decide what ships. The team gets more done without lowering the bar.
I also lead the team: setting direction, hiring, and making it easier for people to do their best work. I still pick up tickets, because a design decision is only as good as your feel for the code it lands in. Shipping alongside the team keeps my judgement honest, shows me where the real friction is, and means I never ask anyone to work in a way I wouldn’t.
NowAs of November 2026
-
Cloud cost optimisation
Rightsizing compute, moving workloads to ARM-based instances and retiring idle resources, cutting spend without giving up headroom.
-
Performance where users feel it
CDN caching and image-delivery tuning, plus query, timeout and payload work on hot paths to bring latency down.
-
Event-driven, scalable services
Queue-backed integrations with retries and dead-letter handling, so traffic spikes are absorbed instead of dropped.
-
Zero-downtime re-platforming
Migrating services to containers one at a time, with replicated data, traffic shifted at the edge and a rollback path at every step.
-
AI in the delivery pipeline
Ticket-driven automation where an LLM coding agent opens a draft pull request in minutes, and ticket and PR state stay in sync through review and release.
-
Observability and hardening
Alerts tuned to sustained failure instead of noise, tighter secrets and access, and origins reachable only through the edge.
What I do
Architecture
Designing systems that are simple to run and easy to change.
Cloud platforms
Running infrastructure as code that is reliable, secure and repeatable.
Modernisation
Moving and upgrading systems without stopping the business.
AI in engineering
Making AI agents a dependable part of how teams build software.
Data and reliability
Keeping data safe, systems observable, and recovery tested.
Engineering leadership
Building teams, habits and delivery pipelines that ship safely.
ExperienceRoles and companies are on LinkedIn.
- 14years
- Building softwareSince 2012, across front end, back end and infrastructure
- 11years
- Full-stack and platform engineeringIn Bangkok, since 2015
- 6years
- Leading engineering teamsFrom team lead to Director of Engineering
StackWhat I work with day to day
- Cloud
- AWS: containers, serverless and CDN
- Infrastructure as code
- Terraform and Docker
- Languages
- PHP, Node.js, TypeScript and Python
- Data
- Relational databases, caching and message queues
- AI
- Coding agents, MCP and LLM APIs
- Delivery
- CI/CD, automated testing and observability
- Security
- Secrets management, least-privilege access and edge protection
Contact
Write to me.
Email works best, and I read everything that comes in. I’m always happy to talk about cloud cost and performance, scalable architecture, AI in the delivery pipeline, or leading engineering teams.