Open Category Submission

859 – Bilden Full Stack AI Education Solution

Bilden delivers a full-stack AI education solution by integrating a standards-aligned curriculum, structured teacher training, and a multi-agent AI platform into a unified system that addresses the core barriers to AI adoption in K-12 schools in China. Our 16-lesson AI Literacy Curriculum, adapted from MIT’s Day of AI and aligned with UNESCO frameworks and China’s universal AI literacy 2030 goals, builds foundational technical knowledge, critical thinking, and ethical awareness in students. The structured Teacher Empowerment Program develops AI fluency through a seven-level growth model, supporting educators from basic AI usage to advanced AI-driven instruction and research outputs. At the technological core, the EduPilot Platform streamlines lesson planning, feedback, grading, and communication through a privacy-first, dual SaaS/on-premise deployment, powered by multi-LLM orchestration and education-specific fine-tuning. Pilot deployments across schools in China have shown transformative outcomes: tripled student interest in AI careers, doubled teacher self-efficacy, and multiple school-level innovation recognitions. With modular, scalable architecture and a strong emphasis on responsible AI use, Bilden creates a sustainable, replicable model that not only meets immediate educational needs but also builds future-ready schools equipped for the AI era.

• Bilden-Pitch.pdf

The idea for Bilden’s Full-Stack AI Education Solution originated from a fundamental supply-demand imbalance in AI education. On the demand side, government-led policies—such as China’s national strategy for universal AI literacy by 2030—rapidly created a top-down imperative for integrating AI into K–12 classrooms. International frameworks, including UNESCO’s, further reinforced the urgency for building global AI competencies. However, the supply side struggled to keep pace: most schools lacked coherent and modular curriculum resources, teachers were largely unprepared to teach AI concepts, and technical infrastructures were either absent or raised concerns over data privacy and usability.

We hypothesized that bridging this implementation gap would require a full-stack approach. Piecemeal interventions—like offering only curriculum toolkits or isolated workshops—were insufficient. Instead, a scalable and sustainable solution must integrate three critical layers: (1) a modular, standards-aligned AI Literacy Curriculum; (2) a structured, practice-driven Teacher Empowerment Pathway; and (3) a secure, AI-powered EduPilot Platform to assist teachers with planning, grading, and communication tasks.

This unified model is grounded in the day-to-day realities of classrooms, not just abstract policy aspirations. Early pilots in Chinese secondary schools validated the model: students showed increased interest and ethical awareness in AI; teachers reported higher confidence and reduced workload; and schools aligned more effectively with national mandates. We believe that only a teacher-centered, full-stack AI education solution can close the implementation gap and truly prepare the next generation for an AI-powered future.

Stream 3

The Bilden Full-Stack AI Education Solution is built on a three-tier modular architecture, combining (1) AI Literacy Curriculum Delivery, (2) a structured Teacher Empowerment Pathway, and (3) AI-Enabled Platform Services.
At its core, the EduPilot platform employs a Multi-LLM Dispatch engine that intelligently routes teaching tasks—such as lesson planning, grading, and communication—to fine-tuned large language models optimized through lightweight education-specific LoRA techniques. Teachers interact through an intuitive agentic AI interface, which abstracts backend complexity and allows seamless use without technical expertise. Innovative features include a unified API gateway for seamless integration of curriculum, training, and AI services, and the introduction of a Model Context Protocol (MCP) that dynamically packages user roles, task types, and curriculum standards for context-aware model responses. Retrieval-Augmented Generation (RAG) modules further enrich outputs with real-time curriculum-aligned references, ensuring outputs are accurate and pedagogically meaningful. Together, these technologies enable sophisticated multi-model orchestration while remaining invisible to users, making advanced AI accessible to all educators.
The design and development timeline included curriculum localization in Q2 2024, prototype completion of the Multi-LLM Dispatch engine and LoRA fine-tuning in Q3 2024, pilot deployments across secondary schools in Q4 2024, and the EduPilot 1.0 release in Q2 2025. Performance metrics are structured across (1) student outcomes (AI knowledge growth, career interest increases), (2) teacher outcomes (self-efficacy in AI-integrated teaching, platform adoption rates), and (3) system outcomes (task automation rate, platform reliability, latency, and data privacy compliance).
Each core function—curriculum delivery, teacher empowerment, and task automation—is directly enabled by targeted technologies, with continuous progress tracked through platform analytics, user feedback, and system performance benchmarks, ensuring scalability, sustainability, and long-term global impact.

The Bilden Full-Stack AI Education Solution offers an innovative and creative response to one of the most urgent challenges in education today: how to meaningfully integrate artificial intelligence into everyday classroom practice without overwhelming non-technical educators. Unlike traditional solutions that focus narrowly on either content or tools, Bilden adopts a full-stack approach that unifies curriculum, teacher development, and AI-driven services into a seamless, scalable system. Technologically, the platform introduces a Multi-LLM Dispatch architecture, dynamically routing teaching tasks across multiple fine-tuned large language models, each optimized for specific educational outputs through lightweight LoRA techniques. The incorporation of the Model Context Protocol (MCP) ensures that user roles, task types, and curriculum standards are systematically embedded into AI interactions, enabling outputs that are highly task-specific and pedagogically meaningful.
Creativity is further reflected in the agentic AI interface design, which abstracts complex backend orchestration—prompt engineering, model dispatching, context handling—into simple, intuitive workflows, allowing teachers to interact naturally with AI assistants without technical barriers. Retrieval-Augmented Generation (RAG) modules enrich AI outputs with real-time curriculum-aligned references, ensuring that responses are both current and educationally grounded. These innovations collectively lower the adoption threshold, increase the precision of AI assistance, and embed ethical and privacy safeguards into everyday use. By solving the implementation gap between AI education policy and real classroom practice, Bilden’s solution enables teachers and students alike to fully leverage AI’s transformative potential, fostering a more inclusive, responsible, and future-ready education system.

The Bilden Full-Stack AI Education Solution is built for both technical scalability and commercial sustainability, combining a hybrid architecture, adaptable delivery pathways, and a go-to-market strategy designed to scale impact responsibly and efficiently.

Technically, the modular AI Literacy Curriculum is deployable as either a standardized package or a customizable program aligned to local academic standards. This flexibility allows diverse schools—from high-performing urban institutions to under-resourced rural districts—to integrate AI education effectively. A hybrid deployment model supports both SaaS (cloud-based) delivery for quick onboarding and on-premise deployment powered by federated learning for high-security contexts. To minimize cost and enhance scalability, we leverage lightweight LoRA fine-tuning, model compression, and task-aware multi-LLM dispatching to reduce computational burden while maintaining performance. A unified API gateway ensures easy platform integration with minimal technical friction.

Commercially, Bilden follows a phased rollout strategy: starting with flagship urban schools as pilots, expanding through regional partnerships endorsed by local education authorities, and scaling up via district- or province-level procurement frameworks. Revenue is generated through a flexible licensing model bundling curriculum, platform access, and teacher training—modularly priced by deployment size and delivery mode. This ensures affordability for schools while supporting sustainable growth. For inclusion, we offer subsidized packages for underserved regions, aligning with public education goals.

Sustainability is embedded in both technology and operations: energy-efficient AI model design and edge computing reduce cloud dependency, while agentic AI interfaces and data-driven teacher pathways adapt to evolving user needs. Regular curriculum updates, user feedback loops, and automated analytics optimize both engagement and platform evolution.

The Bilden Full-Stack AI Education Solution is designed to address persistent social inequities in education—particularly the digital divide that limits access to emerging technologies like AI for students and teachers in under-resourced communities. Traditional AI education models often require significant financial investment, advanced hardware, expert instructors, and cloud infrastructure—barriers that rural, low-income, and marginalized schools cannot easily overcome.

Bilden democratizes access to AI education by intentionally lowering these barriers. Our modular curriculum can be implemented with only basic internet connectivity and functions on lightweight local devices. The Teacher Empowerment Pathway is designed for non-specialist educators, enabling schools without AI-trained faculty to still deliver high-quality instruction. The hybrid deployment model supports both cloud-based and on-premise (federated learning) options, allowing schools to protect student data without bearing the cost of enterprise-level cloud services. By addressing infrastructure, training, and security concerns simultaneously, Bilden provides a complete, scalable solution that even the most resource-constrained schools can adopt.

Social impact is further amplified through our inclusive and ethical approach to content design. The curriculum embeds core themes of AI ethics, fairness, and societal responsibility—ensuring students develop not only technical competence but also critical thinking about AI’s role in society. Our platform fosters equity of participation by enabling all teachers—regardless of technical background—to integrate AI into their classrooms through agentic AI interfaces and hands-on support.

Impact is measured through student engagement and aspiration metrics, teacher confidence and usage indicators, adoption in underserved regions, and ongoing qualitative feedback from users. Responsiveness to local needs is ensured via region-specific content, customization options, and iterative curriculum updates based on real-world classroom feedback.

In advancing both access and quality, Bilden makes a meaningful, measurable contribution to building a more equitable and socially responsible future in AI education.

Personal Information Collection Statement (PICS):
1. The personal data collected in this form will be used for activity-organizing, record keeping and reporting only. The collected personal data will be purged within 6 years after the event.
2. Please note that it is obligatory to provide the personal data required.
3. Your personal data collected will be kept by the LTTC and will not be transferred to outside parties.
4. You have the right to request access to and correction of information held by us about you. If you wish to access or correct your personal data, please contact our staff at lttc@eduhk.hk.
5. The University’s Privacy Policy Statement can be access at https://www.eduhk.hk/en/privacy-policy.

• I have read and agree to the competition rules and privacy policy.