Sbir phase II: fostering growth in trusting relationships between students and educators with natural language processing and machine learning technologies

The National Science Foundation awarded Sown To Grow, Inc. a $1.25 million cooperative agreement on August 1, 2026, under the NSF Technology, Innovation, and Partnerships SBIR Phase II program to develop artificial intelligence-powered tools for measuring and supporting student-teacher relationships in K-12 educational settings.

The recipient is developing a software-as-a-service platform that uses natural language processing and machine learning to make the quality of student-teacher relationships visible, measurable, and actionable at scale. Phase II deliverables include productizing Phase I model outputs into three user-facing tools: student nudges, teacher prioritization features, and administrator dashboards. The company is strengthening technical infrastructure by expanding machine learning models to handle dynamic educational contexts. The proprietary foundation rests on a dataset of millions of real student reflections and teacher responses, creating competitive advantage through models competitors cannot easily replicate.

Phase I, completed prior to this Phase II award, validated three rubric-based classification frameworks demonstrating technical feasibility. This Phase II work runs through July 31, 2028, with performance located in Piedmont, California.

Abstract

The broader/commercial impact of this sbir phase II project is to improve student well-being, academic outcomes, and educator effectiveness by making the quality of student-teacher relationships visible, measurable, and actionable at scale. Strong relationships between students and educators are among the most powerful predictors of student outcomes, yet many schools today lack practical tools to assess or support these relationships in real time at scale. This project addresses that GAP by developing artificial intelligence-powered features that help teachers respond more meaningfully to students, help students deepen their reflective practice, and help school administrators identify where relationships are thriving or need support. The technology is delivered through a software-as-a-service platform. The platform's competitive advantage lies in its proprietary dataset of millions of real student reflections and teacher responses, which underpins machine learning models that no competitor can easily replicate. This small business innovation research (sbir) phase II project advances the application of natural language processing (NLP) and machine learning (ML) to measure and support the quality of student-teacher relationships in k-12 educational settings. Despite the well-documented importance of these relationships, scalable methods for assessing their depth have been largely absent. Phase i established the technical feasibility of this approach by developing and validating three rubric-based classification frameworks. Phase II builds upon this foundation across three research objectives: productizing phase i model outputs into user-facing tools including student nudges, teacher prioritization features, and administrator dashboards; strengthening the technical infrastructure by expanding models to handle dynamic student populations, improving model robustness, and transitioning to real-time inference; and embedding personalized, context-aware teacher coaching directly into educator workflows. Anticipated outcomes include improved quality of student reflections and teacher responses, a robust real-time AI infrastructure, and a scalable model. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the foundation's intellectual merit and broader impacts review criteria.- subawards are not planned for this award.