Mobile Platform Engineer · Mekari
Rehan Choirul
I make mobile builds invisible — the CI/CD, infrastructure, and tooling 20 developers ship on, built mostly by pairing with AI.
I started as an iOS engineer, shipping Swift features to real users. Then I kept following the friction downward — the flaky builds, the release bottlenecks, the pipelines held together with hope — and found that’s where I do my best work. Today I run mobile CI/CD at Mekari for a team of 20: Mac minis, Jenkins, monitoring, and a pile of automation. I don’t really write iOS anymore — the AI era pulled me fully into platform work, and I build faster now than I ever did by hand. The agent writes the draft; the judgment is mine. (I also name my tools after One Piece islands. Take that as a warning.)
Rehan Choirul
Mobile Platform Engineer
Mekari · Jakarta, Indonesia
How I got here
I followed the friction down the stack — from shipping app features, to the platform underneath them, to building the whole thing with AI.
- 01
Shipped iOS product
Built and shipped Swift features to real users — and learned what “done” actually costs in production.
iOSSwiftProduct Engineering - 02
Moved down the stack
Led the migration off Bitrise to in-house infra on Mac minis, and became the platform the whole mobile team’s builds run on.
Mobile PlatformCI/CDInfrastructure - 03
Went AI-native
Now I architect and build tooling — dashboards, bots, CLIs — at AI speed, and keep the engineering judgment human.
AI-AssistedAutomationDeveloper Tooling
Capabilities
Five areas I work across — from the release infrastructure up to the AI that helps me build it.
Platform House
Mobile Platform Engineering
I keep the platform under 20 mobile devs healthy: build infrastructure, Mac minis, monitoring, and the release workflows shipping depends on.
Tooling Workshop
Developer Tooling
CLIs, dashboards, and bots that all 20 devs actually reach for every day — not tools that ship and rot.
Release Tower
CI/CD and Release Systems
Build queues, runner capacity, flaky-failure fixes, and the surfaces that make pipeline state obvious to everyone.
Automation Room
Internal Automation
The boring middle of engineering work: health checks, handoffs, and the shell/Python glue that keeps the machines running.
Agent Garage
AI-Assisted Engineering
I build real, team-used tooling by pairing with coding agents — at roughly 3x my old speed. The agent drafts; I own the judgment.
What I believe
01
The best pipeline is invisible
If developers are thinking about CI, it’s already broken. My job is to make builds something nobody has to notice.
02
AI made me a 3x builder — not a 3x thinker
The leverage is real and I lean on it hard. But the agent writes the draft; the engineering judgment still has to be mine.
03
Ship smaller, validate faster
I once over-built AI features into a dashboard nobody used — the team preferred their own agents. Now I ship the smallest useful thing and let real use pull the rest.
Project Board
Real systems I’ve designed and shipped: tooling, build pipelines, shared libraries, and experiments.
Ohara
Mekari’s mobile engineering command center — build health, pipeline orchestration, release trains, and crash triage in one internal platform.
- Opened daily by all 20 mobile developers — the team’s source of truth
- Turns raw build logs into root-cause categories with AI
- Consolidated build metrics, release planning, and crash triage into one app
Franky CLI
A single Go binary that runs the entire mobile build–test–scan–deploy pipeline for iOS, Android, and Flutter.
- One command surface for build, test, scan, and deploy across three platforms
- Bootstraps a full mobile dev environment from scratch (`doctor --setup`)
- Small static binary with self-update — chosen partly for AI-maintainability
Morgans
A Google Chat bot that lets engineers self-subscribe to exactly the CI/CD and crash alerts they care about.
- Self-service topic subscriptions with per-team severity/platform filters
- Routes build, release, crash, and security events from 19+ pipelines
- Replaced undifferentiated Jenkins spam with signal people opt into
MekariFlag
An internal iOS feature-flag wrapper that standardizes flag usage and provider integration.
- Standardized how flags are read across the app
- Made flag logic mockable and testable
- Swappable provider behind one interface
Agentic Engineering Experiments
How I actually build now — pairing with coding agents to go from task context to a reviewable pull request.
- Turns task context into a reviewable pull request
- Keeps judgment and review with a human
- Removes repetitive context-gathering and boilerplate
Writing
Notes on mobile platform engineering, internal tooling, CI/CD, and agentic workflows.
I Stopped Writing iOS and Started Building the Platform
How the AI era quietly moved me from shipping Swift features to running the CI/CD everything ships on.
Vibecoding Isn’t a Dirty Word
Building real, team-used tools with AI — and why the craft moved from typing to judgment.
Why Mobile Platform Engineering Matters
A short framing note on why mobile teams eventually need platform thinking, not just app feature delivery.
Explore the same portfolio as a 2D world
Game Mode is a larger top-down map with keyboard movement, click-to-move, room labels, and direct project panels.