AI Impact Summit

Global AI Impact Commons

Vinoba

India flag

India

Education

High

Implementing Organisation

Open Links Foundation

India, Maharashtra, Pune

Civil Society

Implementing Point of Contact

Sowmya Srinivasan

Partnerships Manager

Contributor of the Impact Story

Carnegie India

Year of implementation

2025

Problem statement

India faces a persistent learning crisis. While 95 percent of primary school-aged children are enrolled, most perform poorly when evaluated for age-adjusted literacy and numeracy. More than half of children at late primary age are not proficient in reading (World Bank, 2024), and even after 10-12 years of formal schooling, students have limited exposure to vocational education and 21st-century skills. This is despite the government spending about USD 100 billion annually (₹55,000 per student per year) and operating 66 percent of schools in the country, where over 130 million children study – primarily in rural areas. Teachers in these schools are highly qualified, recruited through rigorous competitive examinations, and earn three times more than private school teachers (TISS, 2023). With such extensive investment and infrastructure, government schools are key to improving education quality at scale. Systemic challenges persist; teachers, spread over remote rural areas, spend disproportionate time on administrative tasks, have limited tools for connection and peer learning, and lack recognition and professional support. They lack easy access to quality teaching materials and struggle to address mixed learning levels in classrooms. Communication between administrators and schools is fragmented, often relying on unofficial channels like WhatsApp, with little reliable data to measure the effectiveness of programs. Without a centralized system, program implementation suffers, teacher motivation declines, and the quality of student learning remains compromised.

Impact story details

Open Links Foundation (OLF) was founded in 2016 by Sanjay Dalmia, an alumnus of IIT Delhi and IIM Ahmedabad, after 25 years in corporate leadership. OLF's mission is to make government schools aspirational by improving the quality of education through the Vinoba program. Our theory of change rests on two premises: (a) support and motivate teachers, who are key to student outcomes, and (b) improve program implementation efficiency to enable student practice and feedback at scale. Vinoba supports holistic education through academic and life skill programs via a threefold intervention: 1. A vibrant teacher community, by providing a platform for teachers to connect, share, and learn. Teachers post about projects on poetry recitation, storytelling, spoken English, creative writing, environmental activities, etc. This connects over 200,000 teachers in remote rural areas, who add 50,000 posts each month. 2. Grassroots-level academic support for programs like FLN, 10th and 12th practice tests, and scholarship preparation, with support for program management, data analysis, etc. Each district that we operate in has, on average, 2 field staff working directly with administrators and teachers. 3. Easy access to content, through centralized access to circulars, academic content, activities, worksheets, morning assembly resources, career counselling, mental health content, etc. Without Vinoba, this material remains buried in menus and folders, and discovery stays weak. As local context varies vastly within and across states, the program aligns with the district administration to ensure successful implementation. The Vinoba program has achieved remarkable scale, now covering over 63,000 schools with 2 lakh registered teachers and 6,000+ administrative staff across 37 districts in Maharashtra, Chhattisgarh, Madhya Pradesh, and Bihar, impacting 4.6 million students. The program has demonstrated tangible outcomes across academic programs and life skills activities, providing holistic development for students.

AI Technology Used

Natural Language Processing
Speech Recognition

Key Outcomes

Efficiency

Productivity, Access

Reach, Inclusion

Equity, Accuracy

Quality Improvement, User Experience

Satisfaction, Knowledge

Skills Impact

Narrative Outcome

Vinoba provides an AI-powered platform that reduces teacher workload while improving instructional quality. The platform has reached over 200,000 registered teachers across 37 districts, and tens of thousands of students. Lesson planning time has also reduced by approximately 30 percent. As a result, government spending per student using the platform has dropped to a fraction of traditional costs.

Impact Metrics

Annual cost per student saved through automation on digital platform

Baseline Value

Government spent approximately ₹55,000 per student per annum Indian Rupees

Post-Implementation

Average government spending reduced to approximately ₹25 per student per annum Indian Rupees

Internal monitoring·Apr 2024 - Dec 2025

Number of teachers registered on the platform

Baseline Value

1.46 lakh users (March 2025)

Post-Implementation

Number of registered users reached 2.1 lakh users (December 2025) Users

Internal monitoring·Mar 2025 - Dec 2025

District-level deployment coverage

Baseline Value

24 districts with active deployment (March 2025)

Post-Implementation

Expanded to 37 districts Districts

Internal monitoring·Mar 2025 - Dec 2025

Students impacted as beneficiaries across schools

Baseline Value

24 ,000 students across more than 46,000 schools

Post-Implementation

46 ,000 students across more than 63,000 schools

Internal monitoring·Mar 2025 - Dec 2025

Percentage of pilot teachers regularly using the AI-powered platform

Baseline Value

0 % (pre-implementation)

Post-Implementation

100 % adoption among pilot teachers

Independent Evaluation (NuSocia)·Aug 2025 - Dec 2025

Percentage of teachers expressing high confidence in the platform

Baseline Value

Not applicable Percentage of teachers

Post-Implementation

97.2 % of teachers reported high confidence

Independent Evaluation (NuSocia)·Aug 2025 - Dec 2025

Reduction in lesson planning time

Baseline Value

Not applicable Percentage

Post-Implementation

There was approximately 30% reduction in lesson planning time post implementation Percentage

Independent Evaluation (NuSocia)·Aug 2025 - Dec 2025

Teachers integrating Social-Emotional Value (SEV) content

Baseline Value

Not applicable Percentage of teachers

Post-Implementation

97 % of teachers integrated SEV content

Independent Evaluation (NuSocia)·Aug 2025 - Dec 2025

AI search accuracy (Hit Rate @K)

Baseline Value

66.7 % search accuracy pre-implementation

Post-Implementation

90 –95% search accuracy was reached

Internal monitoring

Reduction in response latency

Baseline Value

15 –25 seconds was the time taken pre-implementation

Post-Implementation

Response latency got reduced to 2–4 seconds Seconds

Internal monitoring·Aug 2025 - Dec 2025

Number of OMR-based assessments conducted

Baseline Value

27 tests

Post-Implementation

356 tests were conducted post implementation

Internal monitoring·Apr 2024 - Dec 2025

Teaching–Learning Material and lesson plan views

Baseline Value

Limited access, content scattered across platforms Number of content/Viewership

Post-Implementation

200 ,000+ curated pieces with 7.63 million+ views indicating wide accessibility post implementation

Internal monitoring·Apr 2024 - Dec 2025

Implementation Context

Deployed

Maharashtra (24 districts), Chhattisgarh (11 districts), Madhya Pradesh (1 district), and Bihar (1 district)

Target population includes more than 2,00,000 registered teachers and 4.6 million students. Demographics include rural government school teachers serving socially and economically disadvantaged communities, students from marginalized groups, including scheduled castes, scheduled tribes, and other backward classes.

Key Partnerships

Maharashtra State Council of Educational Research and Training (MoU for 36 districts), State Council of Educational Research and Training Chhattisgarh, over 45 IAS officers across districts, Meta, The Nudge Institute, HSBC, KMIL, EY Foundation, JC Flowers, Persistent Foundation, IndusInd Bank, GE Shipping, Bandhan Bank, Rustomjee, KR Shroff Foundation, Katherine and Kamal Agarwal Family Foundation, IIM Ahmedabad alumni, and Akanksha Foundation (Nagpur NMC)

Replicability & Adaptation

High

1. The adaptations for semantic search and recommendation functionality can be imagined as Horizontal or Vertical 2. Horizontal when we take it to a new geography and adapt it for a new state, which means a new culture and possibly new languages 3. For example, for morning assembly content – we include state-specific festivals, important people, historical facts, etc. 4. For career counseling – include relevant opportunities and industries in their local context 5. The vertical adaptation will take the technology deeper for more new uses for the same users 6. The technology can be adapted for new domains, for example, searching for circulars, post submissions, SCERT content on Mental health, Financial Literacy, gender sensitivity, etc. 7. SCERT has a lot of high-quality content which is buried in menus and sub-menus and Google Drive directories 8. These can be made easily accessible to the teachers using the semantic search functionality 9. For a new domain – we create the (a) mind maps and (b) data sets 10. Each domain has a particular set of data and a particular way in which a user looks at it 11. For example, a lesson plan always has a learning outcome with an opening, activities, assessment, etc. 12. A morning assembly will always start with prayer, theme of the day, word of the day, an inspirational story, etc. 13. These mind maps need to be developed by creating and curating content and developing the appropriate data sets and further these are programmed in the system 14. Triggers are set so that these can be on the user's recommendation as per the context of the user

Supporting Materials

* The data presented is self-reported by the respective organisations. Readers should consult the original sources for further details.