Making Bharat Entrepreneurial | Udhyam Learning Foundation
Udhyam’s vision of instilling entrepreneurial mindsets in students across the schools in the country is multifaceted and focuses on curriculum-led learning-by-doing. This is powered by a curated curriculum, personalized engagement with students and teachers, real-time support, feedback loops to enhance student learning outcomes and teacher training and facilitation. To achieve our eventual goals, we are looking to build smarter versions of AI-powered chatbots with advanced capabilities, assisted by a web application containing Leaderboards, Portfolios, Curated Project Idea Bank and Content Library that aid the students in their learning journey and empower teachers. These capabilities include:
As students engage in hands-on activities throughout different program phases, they often encounter challenges requiring immediate guidance. Our proposed AI mentor buddy offers students and teachers 24/7 real-time support. This AI mentor answers questions about the "what," "why," "how," and "who" of specific tasks, providing instant, contextual assistance. The AI chatbot also leverages a robust database of past business projects and a high-quality curated content library tailored to students' needs, ensuring relevant and actionable advice. This approach both empowers students to overcome obstacles independently and reduces the burden on teachers to provide constant support.
At different phases of the program, the students are encouraged to submit evidence of their actions in the form of videos, pictures, and textual documents. We are proposing to validate and assess these submissions using scientifically designed rubrics and validation criteria. We propose to utilize AI in this process of validation and assessment of submissions and rank them based on the targeted mindsets and competencies.
We also propose to utilize AI to provide constructive, real-time feedback on the submissions based on the same rubric and evaluation criteria described above. The feedback will focus on both areas done well and areas that can be improved, allowing students to refine their projects and get a higher ranking in the final assessment. By taking the students through the plan →
build → learn → refine cycle, we allow the students to refine their skills to make better real-world decisions.
To improve student and teacher engagement and to encourage timely actions for the completion of the different phases of the program, we aim to utilize personalized AI-generated nudges for both students and teachers. These nudges will be delivered through the chatbot at key stages of the program, contextualized to the specific phase they are in. For example, before the program starts, nudges could highlight the value and benefits of participation, while during the program, they could emphasize the skills that the current activity develops, why they matter, and how the task helps the student gain these skills. This contextualized approach will help motivate students and teachers to be more engaged in their projects and activities.
As described above, the GenAI work at Udhyam includes an AI-enabled chatbot as a learning companion and support tool for students and teachers, possessing the key capabilities to provide real-time help, assess and validate submissions, improve submission quality, and nudge students and teachers to enhance engagement.
Udhyam’s AI Mentor buddy is currently in the late-stage of prototyping phase, with select features undergoing active iteration based on structured user feedback. Initial development began in July last year, followed by a soft launch as a passive support tool in one of the state programs to gather baseline data on usability and engagement. Insights from this phase resulted in key improvements in system architecture and AI response quality. Subsequently, we rolled out controlled pilots in two states, reaching ~300 students with a more engagement-focused implementation strategy. These pilots have allowed us to evaluate performance metrics such as CSAT, engagement rates, and qualitative data on user experience. Based on the results of this user research, we have introduced:
We’re currently working on contextualizing the chatbot responses to every states’ curriculum and program content. We’ll soon be starting the process of evaluating current performance and introducing new features like voice input, automated AI-generated nudges through chatbot etc., before our program relaunch in July. This will be a scale-up phase, targeting a broader rollout across 3 to 4 of the Indian states where Udhyam is running its programs.