It Started with a Simple Question
A few years ago, in a small workspace in Bojonegoro, the Codeiva team sat together and asked: what technology could truly change people's lives? The answer did not come from a global trend, but from everyday observation. We saw students traveling long distances to learn, teachers managing dozens of pupils with limited resources, and parents hoping their children would get a decent education.
That is when artificial intelligence entered our conversation, not as a buzzword but as a tool. We imagined AI that could help teachers prepare materials, grade assignments, or offer extra practice tailored to each student's level. Not to replace teachers, but to give them more time for what matters most: guiding students as human beings.
This new chapter became part of Codeiva's journey as a technology company that grew from a smaller town rather than a major city. We believe AI-driven education does not have to originate in Silicon Valley. It can come from a deep understanding of local context, including patchy internet connections and linguistic diversity.
The Learning Culture We Built Inside the Team
Before building education products, we built a learning culture within our own team. Every week, there is a sharing session where members take turns explaining something new they have learned, whether it is a language model, a data processing technique, or an effective teaching method. We call it learning hour, and attendance is expected, not optional.
This culture made us realize that AI in education is not just about algorithms. There are dimensions of ethics, child privacy, and social responsibility that cannot be ignored. So every product idea goes through long discussions, including hard questions like: what happens if the AI misjudges a student? Who is accountable? How do we protect their data?
Through this process, we learned that the best technology is humble. It knows its limits, knows when to ask for human help, and knows its purpose is to empower, not replace. This principle became the foundation when we began developing educational features within Codeiva's family of products.
AI as an Assistant, Not a Substitute for Teachers
In practice, we find AI most useful when positioned as an assistant. For example, a system can help teachers prepare practice questions at varying difficulty levels, group students by understanding, or give initial feedback on written assignments. Teachers still make the final call, but their administrative workload is significantly reduced.
Another example is adaptive learning. By analyzing patterns in student answers, AI can suggest the next most suitable material. Faster students do not have to wait, while those who lag behind receive extra practice. In areas with limited teachers, this approach can be a lifesaver, not a threat.
We also noticed that many students in Indonesia learn with simple devices and unstable connections. So our design principles always account for this: applications must remain useful offline, be data-efficient, and be friendly to non-technical users. Sophisticated technology is pointless if it cannot be accessed.
Real Challenges: Language, Data, and Trust
There are three major challenges we face. First, language. General AI models are typically trained in English, so their grasp of Indonesian, let alone regional languages, is often inadequate. We have to work extra hard to ensure our systems understand local context, including everyday terms used in schools.
Second, data. Educational data is highly sensitive and scattered across many institutions. Collecting enough data to train a model without violating privacy is complicated work that requires collaboration with schools and government. We choose a transparent approach, where users know what data is collected and why.
Third, trust. Many parents and teachers are still skeptical of AI, and that skepticism is reasonable. We do not try to convince them with grand claims, but with real results and simple explanations. When a teacher sees that their time is more efficient and students are better supported, trust grows on its own.
A Mission Bigger Than Any Product
For Codeiva, artificial intelligence in education is part of a bigger mission: making quality technology accessible to smaller regions, not just big cities. We want students in Bojonegoro, Tuban, or small islands to have learning opportunities equal to those in metropolitan areas. AI, if used wisely, can be a bridge toward that.
This mission is also reflected in how we work. We do not chase the most advanced features, but the most helpful ones. We are prouder to see one teacher save two hours of work each day than to see our model score high on international benchmarks without real impact.
This journey is still long, and we do not claim to have all the answers. But we believe the direction is clear: technology must serve people, especially the younger generation who will carry Indonesia forward. That is the story we want to write, one line of code and one student at a time.