Inspiration
Our journey began with a vision for Indonesia Emas 2045. We realized that the leaders of that golden era are sitting in primary school classrooms today. However, the current educational landscape faces a critical "lack of synergy of the school's triangle": students, teachers, and parents.
We were driven by alarming statistics regarding Generation Alpha:
- Primary school students have an average attention span of only 4-6 minutes.
- While they are tech-savvy, only 30 % can utilize technology to solve complex problems.
- Curriculum acceleration causes 40-50 % of students to be "left behind out of the curriculum target".
We built Pastilulus SD because we believe investing in this platform is an "investment on the Foundation of Indonesia Emas 2045," changing mere curiosity into innovation.
What it does
Pastilulus SD acts as an adaptive "intelligence layer" for primary education. It is designed to "Unlock Every Child's Potential" by personalizing the learning experience.
The platform currently focuses on:
- Adaptive Learning Engine: Offering adaptive and interactive exercise modules that adjust to the student's capability.
- Intelligent Learning Assistant: A digital companion designed to guide students through their academic journey.
- Routine Reporting: Providing basic routine reports on Student Learning Progress to keep parents and teachers informed.
How we built it
Our development approach focuses on creating an intelligent machine capable of supporting academic goals. Building on the success of our Proof of Concept (Pastilulus IELTS), we pivoted to the primary school sector.
Technically, we are integrating advanced AI models into our architecture. Our roadmap includes specific milestones to power our intelligent assistant. We also adopted a microservices architecture, specifically developing a dedicated "microservice for Document Generation" to handle the creation of reports and curriculum materials efficiently.
Challenges we ran into
Transitioning from concept to a deployment-ready ecosystem presented significant technical hurdles. Our team is currently working through:
- Database and Application Stability: We are actively working to "fix errors in database and web application" to ensure reliability.
- UI Implementation: Designing distinct, intuitive User Interfaces for three different user personas—Teachers, Parents, and Students—has been a major focus.
Accomplishments that we're proud of
- Proof of Concept Success: Our predecessor, Pastilulus IELTS, "successfully launched for commercial usage and gain real traction," validating our intelligent machine approach.
- Strategic Partnerships: We secured strategic partnerships with established book publishers (Yudistira & Quadra). This allows us to enter channels parents already trust, which "cuts acquisition costs and builds instant credibility".
- Business Validation: We validated a B2B2C business model using value-based pricing, estimated at 15-20 % of monthly tuition fees.
What we learned
We learned that Gen Alpha's knowledge is often "broad but shallow," necessitating tools that encourage deeper problem-solving rather than just rote memorization.
From a business perspective, we discovered the power of the "school's triangle." Solving problems for one stakeholder alone is insufficient; true impact comes from solving the unsolved problems of students, teachers, and parents simultaneously. We also learned that utilizing an existing distribution network via publishers is key to reaching our target of "mid to high tier (class A - B economy) schools".
What's next for Pastilulus SD
We have a clear roadmap to scale from a pilot to regional expansion:
- Pilot Phase: We are deploying the MVP to 2-4 classrooms (approx. 100 students) to gain actual user data.
- Product Launch: We target launching the MVP in Q4 2025 and the Intelligent Learning Assistant v1 in Q2 2026.
- Future Versions (V2 & V3):
- V2: Will introduce a "Teacher-based Customized Curriculum," Smart Evaluation Reports, and Gamification elements like Leaderboards and XP.
- V3: Will implement "Expert-based Talent Mapping" and "Student Psychology Based Report".
- Expansion: We aim to expand to the wider Jabar-Jabodetabek region, targeting minimum 0.4 % of the Serviceable Obtainable Market (SOM).
Built With
- drizzle
- gemini-3-flash-preview
- gemini-3-pro-preview
- neon
- nextjs
- orm
- postgresql
- shadcdn
- tailwindcss
- typescript
- vercel
- vercel-ai-sdk
Log in or sign up for Devpost to join the conversation.