By Dave DeFusco
When Ngoni Shaani enrolled in Professor Jiang Zhou鈥檚 Artificial Intelligence course in the Katz School鈥檚 M.S. in Artificial Intelligence, he had no idea a single class project would ignite one of the most ambitious education technology initiatives in Zimbabwe. The assignment, which involved applying natural language processing (NLP) and semantic similarity models to align content to the Common Core State Standards, was meant to teach technical skills. For Shaani, however, it inspired a vision: what if the same technology could help modernize and safeguard the quality of curriculum materials across his home country?
That idea has grown into a working AI prototype with the potential to reshape Zimbabwe鈥檚 education system. Called the Zimbabwe Curriculum Standards Alignment AI Tool, the prototype represents a fusion of technical ingenuity, lived experience and a deep commitment to educational equity.
鈥淎 special thank you goes to Professor Zhou,鈥 said Shaani. 鈥淗is lecture and project taught us NLP and provided the foundation for applying AI to curriculum standards alignment in a practical, scalable way.鈥
Zimbabwe is currently overhauling its national education framework, shifting to a competency-based curriculum that aims to modernize teaching and learning. The transition is ambitious, yet vulnerable. The market has been flooded with unofficial textbooks and learning materials that are being widely purchased and often fail to meet national standards. Teachers are left guessing, parents are overwhelmed and policymakers lack insight into what students are actually learning.
For Shaani, the problem felt personal. Growing up, he experienced firsthand how unequal access to quality learning tools can shape a child鈥檚 educational trajectory.
鈥淚n primary school, our entire school shared just five computers,鈥 he said. 鈥淲e each got maybe 15 minutes per week, but that small window changed my life. It鈥檚 what sparked my interest in computer science. I know many kids today still face those same limitations.鈥
His own late introduction to technology and the effort it took to catch up shaped his drive to use AI to promote educational equity. Shaani鈥檚 prototype tackles the curriculum-quality crisis head-on. Using advanced NLP techniques, the tool evaluates textbooks, lesson plan and digital resources, and checks how closely they align with official standards from the Ministry of Primary and Secondary Education. Its core features form a powerful ecosystem:
- Open Standards Dataset: A machine-readable library of competencies extracted from decades of curriculum documents, exam papers and syllabi.
- AI Alignment Engine: An automated evaluator that analyzes content, generates embeddings and measures semantic similarity to national standards.
- Web Portal and API: Enabling teachers, publishers and policymakers to upload materials, check alignment scores, generate reports and search standards.
The innovation is especially transformative in a system where curriculum reviews are slow, expensive and highly manual. 鈥淎I agents can help review materials faster,鈥 said Shaani. 鈥淭hen humans can refine the results, which would be true human-AI symbiosis.鈥
If implemented nationwide, the tool could change how Zimbabwe evaluates and produces educational content. Teachers would plan lessons confidently, knowing their materials align with national goals; parents could distinguish high-quality resources from misleading ones; publishers would ensure compliance before releasing textbooks; policymakers would monitor curriculum integrity in real time; and Zimbabwe鈥檚 education sector would gain transparency, accountability and global benchmarking against frameworks like Common Core State Standards. The tool functions like a 24/7 curriculum expert鈥攁n innovation that could reduce review costs, compress timelines and dramatically improve educational quality.
Shaani鈥檚 commitment to education did not begin with this project. He previously founded My Grad Connect, a platform linking students from poor countries to global scholarships and opportunities, reflecting the support that helped bring him to the Katz School. He also developed Rumi, a tool that helps Zimbabwean students find compatible shared accommodation, and he is currently building an offline AI learning platform with another Katz School student, Emmanuel Kasigazi, for communities with limited internet access.
鈥淓verything goes back to limited resources,鈥 said Shaani. 鈥淲hen you understand the barriers firsthand, you build solutions differently.鈥
Creating the AI tool wasn鈥檛 easy. Extracting, cleaning and structuring standards from massive text-based documents was a painstaking process. 鈥淭he dataset was the toughest part,鈥 he said, 鈥渂ut it鈥檚 growing every day.鈥
As the prototype moves toward public testing, Shaani plans to measure its impact through cost reduction, time saved during curriculum reviews, user feedback and engagement metrics from open-source testers. The vision is bold: an evolving, AI-powered system that keeps pace with rapid technological change, which is something traditional curriculum frameworks cannot do.
鈥淚f this tool can touch lives back home, even a fraction of what Professor Zhou鈥檚 course did for me, then it will have achieved its purpose,鈥 he said.
