Subodha LMS case study

Project Overview

This capstone project is about designing a new analytics dashboard for Subodha LMS. Subodha LMS is a learning management system that is used by several organisations involved in providing STEM (Science, Technology, Engineering, and Mathematics) education to children who are blind or visually impaired (BVI).

The goal of this project is to build a dashboard that gives the team managing the platform a much clearer picture of what is happening on the platform. The dashboard should help them understand how the platform is being used, what kind of impact it is creating, and where they need to focus their energy. It should also help the organisation tell a meaningful story about its social impact, which is important when they are meeting with donors, government partners, or funding organisations.
The existing analytics tools used by the team have not been able to fulfil these needs. They either show raw numbers without any useful context, or they were built for commercial businesses and do not fit well with how a non-profit organisation like Vision Empower works. This project aims to design a purpose-built dashboard that solves these problems.

The Problem: Vision Impairment and Education in India

Scale of the Issue

Visual impairment affects a very large number of people around the world. There are approximately 2.2 billion people globally who are visually impaired. India alone accounts for roughly one-third of that total, making it one of the countries with the highest concentration of people with visual impairment.
Despite this scale, access to education for visually impaired children in India remains very limited. Government policies do require schools to provide inclusive education, but the reality on the ground is quite different. According to data from the Census of India conducted in 2011, only 68 percent of BVI children between the ages of 5 and 19 were enrolled in school. Even more alarming, around 20.4 percent of BVI children in that age group had never attended school at all.

The Teaching Infrastructure is Constrained

India has more than 400 special schools that are dedicated to educating students who are blind or visually impaired. These schools collectively employ around 2,000 teachers. An interesting fact is that 60 percent of these teachers are themselves persons with visual impairment. This means the people teaching BVI students often understand the challenges from personal experience.
However, this also means the infrastructure is limited in scale. The number of trained teachers available is not enough to meet the needs of the large population of BVI children across the country. The problem is not just about buildings and classrooms. It is also about the availability of accessible learning materials, trained teachers, and the right tools to support teaching.

The Problem Goes Beyond Physical Limitations

Visual impairment is not just a physical condition. It is also accompanied by significant psychological and social challenges. Children with visual impairment often grow up in environments where they are seen through the lens of their disability, which affects how they see themselves. Societal assumptions that disability defines a person's worth, combined with stigma and stereotyping, compound the difficulties that BVI children already face in accessing education.
This means that addressing the problem requires more than just providing textbooks in braille. It requires building a system that is empowering, that treats BVI children as capable learners, and that gives teachers the tools and knowledge they need to create meaningful learning experiences.

STEM Education Remains Largely Inaccessible

Even though there is broad agreement that all children deserve a good education, STEM subjects remain particularly hard to access for visually impaired students. Most STEM learning materials involve visual representations such as diagrams, graphs, and models. Adapting these into accessible formats is a complex and resource-intensive process.
Employment also remains a challenge even for BVI individuals who manage to complete their education. The world of work has not caught up with the needs of BVI professionals, and suitable opportunities remain limited. This makes it even more important to invest in quality education from an early age, so that BVI children can develop the skills and confidence they will need throughout their lives.

About Vision Empower

Vision Empower is a non-profit trust that was established with the goal of making inclusive education possible for students who are blind or visually impaired. The organisation was incubated at the International Institute of Information Technology Bangalore (IIIT Bangalore) and works in partnership with a range of organisations to develop tools and materials that help BVI children learn.

Vision Empower works on several fronts. It conducts foundational research to understand the specific needs of BVI students and teachers. It develops accessible educational content, including braille textbooks and tactile diagrams that allow students to feel and understand scientific concepts. It also runs teacher training programmes that equip educators with the skills they need to teach BVI children effectively.

The organisation has received support from a number of prominent sponsors and partner organisations, including Microsoft, Wipro, the Cognizant Foundation India, HP, Fidelity Investments, and Elektrobit. These partnerships reflect the credibility and the importance of the work that Vision Empower is doing.

However, Vision Empower recognised early on that creating good content was only part of the solution. Without a reliable and accessible way to distribute that content to teachers and students across different schools and states, the work would remain fragmented and limited in reach. This is what led to the creation of Subodha, the learning management system at the centre of this project.

About Subodha LMS

Subodha is a learning management system developed by Vision Empower. The name itself reflects the organisation's commitment to making learning accessible and meaningful. Subodha is built on top of openEdX, which is an open-source teaching and learning platform that was originally developed by the Massachusetts Institute of Technology (MIT). This foundation gives Subodha a solid technical base while allowing Vision Empower to customise it for their specific needs.

Subodha contains courses across a range of subjects and educational boards, catering to students at different class levels. The primary type of content available on the platform is called Teacher Instruction Kits, or TIKs. These are carefully designed learning packages that are created for teachers who are working with visually impaired children. A TIK gives a teacher everything they need to teach a particular concept in a way that works for BVI students, including step-by-step instructions, accessible materials, and guidance on how to use tactile tools.

Subodha functions as both a learning management system and a content management system. The platform serves two main types of users: teachers at partner schools who use it to access TIKs and prepare for their lessons, and Vision Empower employees who use it for internal training courses. Students are generally not direct users of Subodha because most BVI students in schools access digital resources through computer labs or dedicated assistive technology devices, and smartphones are not commonly available to BVI children in most states.

In the last financial year, Subodha added new features including a voice search function and a certification module. These were added as part of the A4IT (Assistive Technology and Artificial Intelligence) initiative. Voice search is especially important for BVI users who find it difficult to type on standard keyboards.

Partner organisations that work with Vision Empower and Subodha include Xavier's Resource Center for the Visually Challenged, Manprax, NIEPVD (National Institute for Empowerment of Persons with Visual Disabilities), and Divyangjan.

The Capstone Project

Problem Statement

The formal problem statement for this capstone project is as follows: to design the analytics dashboard for Subodha LMS. The new dashboard must provide a better picture to the team that manages the platform, enabling better prioritisation of features and helping them build a compelling story around social impact that supports the cause of the Vision Empower Trust.

In plain terms, the task is to design a dashboard that actually works for the people who use it. The current tools are not doing the job. This project starts from scratch with user research, understands what different stakeholders need, and then designs a solution that addresses those needs.

Why Dashboards Matter for an Organisation Like This

It is worth pausing to ask why an analytics dashboard is so important for a non-profit organisation like Vision Empower. The reasons are specific and significant.

First, the beneficiaries of the platform cannot advocate for themselves in the data. Visually impaired students in rural schools do not have a way to flag problems or show dissatisfaction with the platform. Unlike users of a commercial product, they will not write reviews or send complaints. If the platform is not working for them, they simply stop using it quietly. Without a proper dashboard, this kind of disengagement is invisible to the team that manages the platform. The dashboard is the only tool that can make the experience of these students visible to the people who have the power to change things.

Second, non-profit organisations do not have market mechanisms to tell them when they are failing. A commercial product will see its revenue fall if it is not serving its users well. A non-profit can continue for years while silently underserving its mission, because there is no financial signal to trigger a course correction. A good analytics dashboard creates the feedback loop that non-profits otherwise lack.

Third, the entire impact of Subodha depends on whether teachers are genuinely engaging with the TIKs they are assigned, not just logging in and leaving. Without the right metrics, the team has no way to distinguish between a teacher who has logged in once and a teacher who has deeply engaged with multiple TIKs and applied the learning in their classroom. This distinction is crucial, but it is completely invisible without the right data.

Fourth, funders, donors, and government partners all need evidence that the organisation is delivering on its promises. Every grant renewal, every funding conversation, and every state-level partnership requires Vision Empower to demonstrate what has changed, for whom, and where. A well-designed dashboard converts the daily work of the team into the kind of evidence that funders need to continue their support.

Fifth, organisations that can identify and respond to problems quickly are more effective over time. An organisation that sees a problem in days rather than months can take action sooner, waste less effort, and improve faster. Over a period of five years, this difference in speed compounds into a significant gap between a platform that works and one that does not.

Problems with the Current Analytics Setup

A History of Switching Platforms

Subodha has gone through three different analytics platforms over the course of its existence. The first was the analytics feature built into openEdX itself. After that, the team moved to Google Analytics. Most recently, they switched to Apache Superset, which is an open-source analytics platform that draws its data from openEdX.

Each switch was driven by dissatisfaction with the previous tool. The core problem was consistent across all three platforms: they were not designed with an organisation like Vision Empower in mind. Google Analytics, in particular, is built for commercial websites and e-commerce. Its metrics are designed to track revenue, acquisition, and customer behaviour. These concepts simply do not map well onto the work that Vision Empower is trying to do.

Superset was chosen partly because it was simpler than Google Analytics and more accessible to non-technical users. It offered a small set of options and was self-contained. However, even Superset fell short because it still failed to show a proper user journey and could not translate the data it collected into actionable insights.

Specific Problems Identified

After reviewing the existing Superset dashboard, several clear problems were identified:
• The dashboard presents raw data without any guiding narrative. There is no story being told. Numbers appear on screen but they do not connect to any meaningful interpretation of what is happening on the platform.
• The metrics tracked are access counts rather than outcomes. The dashboard tells you how many people logged in, but not whether those people actually learned anything or changed how they teach.
• There is no aggregation, no comparison between time periods, and no trend visualisations. You cannot see whether things are getting better or worse over time, or compare performance across different schools or states.
• Completion rate figures are shown without any context. A high completion rate might look good but could be misleading if the content is very short or if the same teacher is completing the same TIK multiple times.
• There are no actionable insights for management. The dashboard does not help anyone decide what to do next.

Beyond the technical limitations, there was also a deeper conceptual problem. The dashboard was telling a revenue story rather than an impact story. It was counting things rather than measuring meaningful change. For an organisation that exists to improve the lives of visually impaired children, this is a fundamental mismatch.

Impact on the Organisation

This analytics gap has created real problems for Vision Empower. When the team goes into meetings with investors or sits down with donors to make the case for continued funding, it is very difficult for them to communicate the value of their work in a way that is backed by data. They know that teachers are using Subodha and that students are benefiting, but they cannot prove it clearly or quantify it with confidence.

The team at Subodha believes that a well-designed analytics dashboard will make a significant difference. It will give them the ability to show funders exactly what their money is achieving, help them manage the platform more effectively, and build a stronger case for the importance of the work they are doing.

Research and Analysis

Research Approach

The research for this project involved conducting interviews and meetings with people in different roles who are connected to Subodha and Vision Empower. The goal was to understand what each type of user needs from an analytics dashboard, what questions they are trying to answer, and what decisions they need to make.

The roles included in the research plan were: the Programme Director, the Product Manager for Assistive Technology, Course Creators, Teachers, and an Accessibility Expert with expertise in WCAG (Web Content Accessibility Guidelines).

Findings from Dhruba Narayan (Assistive Technology Specialist)

Dhruba Narayan is an Assistive Technology Specialist and Account Manager at Vision Empower, and previously served as a State Coordinator for the organisation. He works primarily in the field, dealing with strategic and technology-related matters.

From the interview with Dhruba, several important things were learned about how Subodha works and what kinds of data are needed.

On the topic of the platform's ecosystem, Dhruba confirmed that Subodha has two main user groups: teachers at partner schools, and Vision Empower employees who take internal courses. Students are not primary users of the platform because of device restrictions. In most states, BVI children access digital resources through school computer labs or through dedicated assistive technology devices such as the Access Iris. Smartphones are generally not available to BVI students in these settings.

On the topic of analytics history, Dhruba explained that the platform has gone through three generations of analytics tools. Google Analytics came first, followed by Figures (a built-in openEdX feature), and then Apache Superset. Each replacement was driven by dissatisfaction with the previous tool. Superset was chosen because it was simpler and more accessible for non-technical users.

On the topic of metrics, Dhruba outlined the core data points that are most useful to him. These include the number of teacher logins, active user counts, time spent on courses on a
week-over-week basis, course enrolments, certification completions, and induction course completions. He also highlighted voice search as a particularly important metric because voice input was specifically added to the platform to support visually impaired users who struggle with typing. Tracking the volume of voice searches and how that correlates with overall engagement is therefore a meaningful indicator.

Dhruba also raised the question of content impact. The team wants to understand how much time teachers are spending on specific content, such as the NIEPVD training videos, and whether that time investment translates into better outcomes for students. However, he acknowledged that student-level analytics are not currently feasible within the Subodha platform.

Findings from Meera Muthukrishnan (Programme Director)

Meera Muthukrishnan is the Programme Director at Vision Empower. Her perspective was focused on what the dashboard needs to communicate and to whom.

She was clear that the dashboard should lead with outcomes rather than metrics. Data points like the number of logins, enrolments, or time spent on the platform only matter if they connect to a larger story about impact. Did a teacher who completed a TIK go on to teach that concept better in their classroom? Did a child with visual impairment benefit from that teacher's improved preparation? These are the questions that matter, and the dashboard needs to help answer them.

Meera emphasised that the primary audience for this data is not a corporate board looking at revenue. It is grant-makers, donors, and partner organisations who need to understand what has changed as a result of Vision Empower's work, and for whom that change has happened. The language, visualisations, and structure of the dashboard need to reflect this. Revenue-style KPIs are not just unhelpful in this context; they are actively misleading.

She also highlighted the importance of answering the 'so what' question at every level of the dashboard. Every number or visualisation should connect to a meaningful insight. If a metric does not help someone understand whether the organisation is making progress toward its mission, it should not be on the dashboard.

Findings from Chandrashekhar B Korlahalli (Accessibility Consultant)

Chandrashekhar B Korlahalli, also referred to in the project as 'Chandra sir', is an accessibility consultant at Vision Empower. He is both an expert in accessibility standards and a person with visual impairment who uses assistive technology in his daily life. This combination of professional expertise and lived experience makes his input particularly valuable.

One of the most important themes that emerged from the interview with Chandra was the idea that accessibility should be understood as a lived experience rather than a compliance checkbox. His own daily use of tools like NVDA (NonVisual Desktop Access), TalkBack, and screen magnifiers grounds every recommendation he makes in practical reality. Accessibility is not an abstract requirement for him. It is something he navigates every day.

He also raised a broader structural issue: in the Global South, feedback from disabled users is frequently suppressed. This happens because of cultural deference, broken escalation paths, and the absence of enforcement bodies. Disabled users often do not feel empowered to complain, and even when they do, their feedback does not reach the people who can act on it. This is not a personal failing on the part of any individual; it is a systemic problem that is reflected in both physical and digital infrastructure.

On the topic of accessibility standards, Chandra explained that WCAG (Web Content Accessibility Guidelines) is the universal root standard from which national standards in countries like the United States, France, and others are derived. Designing to WCAG effectively satisfies most regional requirements. This simplifies the compliance target significantly.

For screen readers, NVDA is the de-facto standard for vision-impaired users in India who cannot afford the more expensive JAWS screen reader. TalkBack is used on mobile devices. Chandra noted that screen reader announcements can sometimes diverge from what is visually displayed on screen, which creates a confirmed bug pattern that needs to be tested during the design process.

Chandra had already conducted an accessibility audit of the existing Subodha platform before this project began. Most issues had been addressed, but a few remained open. The new dashboard design starts with awareness of this existing design debt.

His most important practical recommendation was to incorporate accessibility review at the wireframe stage, before any code is written. He described this approach as 'shift-left
accessibility', drawing on his experience at HSBC where early-stage review significantly reduced the number of defects in the final product.

He also discussed the importance of communication and collaboration infrastructure. He prefers email over messaging apps like WhatsApp because it creates a traceable record. This practical consideration is itself a design constraint: the working relationship between the designer and the accessibility consultant needs to be structured in a way that works for both parties.

Design Process

Methodology

This project follows the Design Thinking methodology, which is a structured approach to solving problems that puts the user at the centre of every decision. The five stages of Design Thinking are Empathise, Define, Ideate, Prototype, and Test. The process is iterative, meaning that after testing, the team returns to the earlier stages to refine their understanding and improve the design.

In the Empathise stage, the focus is on understanding the needs, frustrations, and goals of the people who will use the product. This is done through interviews, observations, and conversations with stakeholders. In the Define stage, the insights from the research are synthesised into a clear statement of what problem is actually being solved. In the Ideate stage, a wide range of possible solutions are explored before narrowing down to the most promising ideas. In the Prototype stage, early versions of the design are built and made tangible. In the Test stage, these prototypes are put in front of real users to see whether they work as intended and to gather feedback for further refinement.

Tools Used

The following tools were used across different stages of the design process:

Prototyping: Figma, which is a professional design tool widely used for building interactive mockups and prototypes of digital products.

Research Analysis: Claude (an AI model by Anthropic), which was used to help analyse interview transcripts and extract key insights from the research data.

Transcription: TurboScribe, which was used to transcribe audio recordings of interviews into text for analysis.

Competitive Analysis

As part of the research phase, the team reviewed existing analytics dashboards to understand what good design looks like in this space and what can be learned from products that are already on the market. Two key platforms were reviewed.

Google Analytics is a very powerful and widely used analytics platform. It offers extensive data on user behaviour, acquisition channels, engagement, and conversions. However, it is designed for commercial websites and e-commerce businesses. Its entire framing is around acquiring
customers and driving revenue. This makes it a poor fit for a non-profit platform like Subodha, and one of the main reasons the Subodha team moved away from it.

Google Play Console Analytics was also reviewed as an example of a product analytics dashboard that is focused on performance and user satisfaction rather than revenue. It organises data around metrics like user acquisition, ratings and reviews, crash rates, and store performance. While this is also designed for a commercial context, some of its structural choices around presenting KPIs with comparisons to previous periods are useful reference points for the Subodha dashboard design.

Current Design Work and Prototypes

Overall Metrics Screen

The first screen that has been designed and prototyped is the Overall Metrics screen. This screen is designed to serve as the entry point to the dashboard and give the user a clear summary of what is happening across the platform.

The screen opens with a headline impact statement. For example, it might say something like: 'This month, 312 teachers strengthened accessible STEM education for an estimated 6,240 students.' This immediately tells the viewer what the platform has achieved in a language that is meaningful and human, rather than simply showing raw numbers.

Below this headline, a set of key metrics are presented as cards. These include active teachers in the current period, new teacher registrations, certifications earned, and average time spent per TIK session. Each metric is accompanied by a comparison to the previous period, so the user can quickly see whether things are improving or declining.

The screen also includes a section called 'How Impact Flows: Teacher Preparation Pipeline'. This is a visualisation of the stages that a teacher goes through on the platform, from the TIKs being published by content creators, to being assigned to teachers, to being opened, to being completed, and finally to the teachers returning for a second TIK. Each stage shows the count of teachers or TIKs at that stage, and the proportion of the pipeline that has reached that point. This helps the team see where teachers are dropping off and where the platform is working well.

Another section on this screen is called 'Teacher Engagement Depth'. This visualisation shows how many TIKs each teacher has completed, categorised into groups: teachers who have completed five or more TIKs (described as deep engagement), those who have completed two to five TIKs (described as growing), those who have completed one TIK (early stage), and those who have not completed any TIK and may need follow-up.

A reach and accessibility section is also included. This shows the number of states with active teachers, the number of districts reached, the percentage of sessions using screen readers, and the number of voice search queries made in the current week.

Design Principles

Several guiding principles have shaped the design work:
The dashboard tells an impact story rather than a revenue story. Every metric and visualisation is chosen because it helps the viewer understand whether the organisation is moving toward its mission, not because it is a standard business KPI.
Data is spread across different pages for different user roles. Teacher data, student data, and coordinator data are on separate pages. This helps each type of user find the information they need quickly, without having to wade through data that is not relevant to them.
The overall metrics page acts as a summary that gives a clear picture of impact and progress at a glance. Users who need more detail can navigate to the relevant section of the dashboard.
Accessibility is a core consideration throughout. The dashboard is being designed with screen reader compatibility in mind, and accessibility will be tested formally as part of the testing phase.

Further Steps and Project Roadmap

The project is currently in progress. The following steps are planned to bring the project to completion:

1. One further interview will be conducted with a teacher or programme manager who works at the ground level. The goal of this interview is to deeply understand the day-to-day teaching methodology and the specific challenges that teachers face when working with BVI students. This will help ensure that the dashboard reflects the realities of teaching, not just management-level concerns.

2. One interview will be conducted with a course creator to understand how TIKs are made, what decisions go into creating them, and what data a course creator would need to evaluate whether their content is working.

3. Based on the additional interviews and feedback gathered, prototyping will continue for the other screens of the dashboard. This includes screens for User Metrics, Programme Metrics, Course Metrics, and Certificate Metrics.

4. Once the prototypes are ready, usability testing will be conducted with programme managers and directors. This testing will evaluate whether the dashboard is easy to use, whether it addresses the project objectives, and whether it meets accessibility compliance requirements.
   The prototype developed so far is available on Figma and can be accessed at the link provided in the project files. A project log is also being maintained to document all decisions, findings, and changes made throughout the project lifecycle.

   
   

   Literature Review

   
   

   The following sources were reviewed as part of the research and contextualisation for this project:

   1. Vision Empower Trust (2024). Subodha LMS Submission Paper. This document, published by Vision Empower Trust, provides background on the Subodha platform, its goals, and its development history. It was used to understand the technical and organisational context in which the dashboard will operate.

   2. Parthasarathy, B., Dey, S., and Gupta, P. (2021). Overcoming wicked problems and institutional voids for social innovation: University-NGO partnerships in the Global South. Technological Forecasting and Social Change, Volume 173, Article 121104. This academic paper examines how partnerships between universities and NGOs can be used to drive social innovation, particularly in contexts where existing institutions are not able to address complex problems. It provides relevant background on the kind of institutional dynamics within which Vision Empower operates.

   3. Vision Empower Annual Report, 2024-25. This is the most recent annual report published by Vision Empower. It contains data on the organisation's reach, activities, and impact during the 2024-25 financial year, including information on the number of teachers reached, students supported, and partnerships maintained. It was used to ground the project in accurate, up-to-date information about the organisation's scale and operations.