Scaling Access to High Quality Contextualized Teaching Materials for 3m+ Indonesian Teachers
In 2022, I joined Indonesia’s Ministry of Education as a Senior Product Manager for Platform Merdeka Mengajar (PMM), just as the country was rolling out its most ambitious curriculum reform in decades. Kurikulum Merdeka (translated as the “Emancipation Curriculum”) redefined the role of the teacher — from executing fixed lesson scripts to designing instruction responsive to their students' needs and classroom realities.
CHALLENGE
For many, this shift was disorienting. Years of working within rigid, centralized systems had not prepared teachers to make pedagogical decisions or adapt materials to diverse local contexts. Yet under Kurikulum Merdeka, that level of contextualization was foundational to its mission of student-centered, differentiated learning.
To support teachers through this transition, the platform’s Perangkat Ajar feature served as a national library of curriculum-aligned lesson plans, teaching materials, and assessments. But while demand for content was high, the supply side wasn’t built to scale.
The content team initially partnered with education institutions and manually selected expert teachers to contribute materials. Content was passed through a centralized moderation workflow that was always under strain. Review backlogs were constant. The pipeline simply couldn’t keep pace with the scale of what teachers needed—especially across Indonesia’s geographically and culturally diverse regions.
CONTRIBUTION
Drawing from my prior experience building a creator-driven content ecosystem at Noice, I set out to reimagine the contribution model behind Perangkat Ajar. But unlike a consumer content platform, PMM came with far higher stakes:
Every piece of content carried the imprimatur of the Ministry of Education.
Errors risked reinforcing misconceptions in classrooms.
With those constraints in mind, I designed and led the development of a scalable, peer-powered contribution ecosystem anchored in trust, context, and quality.
🪜 Teacher Growth Pathways
We introduced a structured “leveling-up” system that allowed teachers to grow within the platform (user —> contributor —> reviewer):
Contributor Onboarding: Modular training and clear submission guidelines gave teachers the tools to start contributing confidently.
Built-in Feedback: Contributors received visibility into how their content was used, plus feedback from reviewers, turning submission into a professional development loop.
Peer Reviewer Pathway: Experienced contributors could “graduate” into reviewer roles, applying rubrics and supporting quality assurance while lightening the load on central teams.
⚙️ Moderation System Redesign
Tag-based triage workflows prioritized high-quality submissions
Quality tiers and escalation paths helped focus review effort where it mattered
Feedback tools enabled more constructive, transparent reviewer-contributor interactions
Decentralized institutional contribution oversight, allowing trusted institutions to manage their own contributor onboarding and internal review process
The result: increased throughput, reduced operational overhead, and a more responsive contributor experience.
❋❋ The edtech landscape has shifted dramatically in recent years. AI-powered lesson plan generators built on large language models can now produce custom instructional materials in seconds, tailored to any topic, age group, or pedagogical style. In contrast, the system we built was human-powered, intentionally slow in some places, and deeply invested in teacher development as much as content output. A machine might generate a decent worksheet, but it can’t model how a teacher in rural Kalimantan adapts a lesson for a mixed-grade class, or how a Bahasa Indonesia teacher embeds local proverbs into a reading module to deepen student connection. That kind of pedagogical nuance lives in community, not code. That said, the future isn’t binary. The infrastructure we built (peer-contribution pathways, distributed moderation, feedback loops) could be the foundation for a hybrid model: one where AI helps seed ideas, accelerate iteration, or fill content gaps, but human educators still serve as curators, translators, and context-builders. ❋❋
