Governed Meta Ads
Building safe, agent-accessible workflows for one of the world's most powerful marketing platforms.
Marketing engineering for AI agents.
Syed Akmal Syahmi / Founder, Blytz Ventures / Kuala Lumpur
I build governed infrastructure that helps autonomous agents understand business context, operate marketing tools safely, and learn from outcomes.
Currently building Blytz·AI agents / MCP / Meta Ads / Cloudflare
27public repositories
28+Cloudflare Workers
12subagents in my dev setup
KLKuala Lumpur, Malaysia
Building the interface between business and automation.
Marketing Engineering is the discipline I'm exploring at Blytz: give agents the context, tools, and feedback loops to do useful work without losing control.
Reusable agent skills
Turning hard-won marketing context into repeatable skills that autonomous agents can actually use.
Learning workflows
Designing systems that run experiments, read outcomes, and improve the next decision.
Blytz as execution layer
Connecting context, action, and feedback in edge-native infrastructure for AI-driven marketing.
Selected work from GitHub.
A current slice of the public work around Blytz, EmDash, MCP, and agent infrastructure.
Latest build
blytzintel-skills
Agent skills for using Blytz Intelligence MCP.
Commerce distribution
emdash-ecommerce-native
Native EmDash ecommerce distribution with Commerce Core.
Core contracts
emdash-commerce-core
Provider-neutral Commerce Core and Contracts for EmDash.
Payments integration
chip-for-emdash
Accept CHIP payments — FPX, e-wallet, card, and DuitNow QR — on any EmDash site via hosted checkout.
Tools for feedback loops.
I prefer systems that are useful quickly, economical to run, and understandable when an agent makes a surprising decision.
Build principle
Keep the runtime close to the work.
StackCurrent
Agent layerMCP · agent skills · governed tool access
MarketingMeta Ads · experiments · outcome loops
RuntimeCloudflare Workers · D1 · edge infrastructure
BuildGo · TypeScript · AI-assisted delivery
AI agents
Agentic systems as a product surface.
MCP
Context-rich interfaces for external systems.
Cloudflare Workers
Edge execution for low-latency operations.
Go + TypeScript
Pragmatic systems from service to surface.
Build something useful.
Explore the work, follow the build, or start a conversation about AI-native systems.