At our core, we believe that true impact goes beyond simply providing a product or service—it’s about ensuring that organizations can leverage technology to its fullest potential. Our focus on sustainability drives us to build long-term solutions rather than short-term fixes, ensuring that NGOs not only adopt the platform but also integrate it meaningfully into their work. We are impact-driven, focusing on measurable outcomes that enhance how these organizations operate and engage with their communities. Our commitment to excellence means that our success is tied to the success of the organizations we support.
This philosophy comes to life through the NGO Sprint, where social sector-focused organizations gather in a single space, exchanging ideas over walks, presentations, workshops, meals, and coffees. These conversations create a space for NGOs to find common ground, dive deep into challenges—both tech and non-tech—and for the Glific team to truly connect with the people behind the usernames on Discord and email addresses using the platform.
Understanding the stories behind each chat, knowing how a simple WhatsApp message can unlock opportunities in education, careers, health, or essential benefits, is what fuels our motivation to keep pushing forward, ensuring that NGOs don’t just use the platform but truly excel with it.
Glific Makers Gathering: Jan 2025
Based on the sensing (what we hear and provide dedicated help to NGOs), the Glific sprint, this for the month of January was decided to focus on 3 major themes.
- Help NGOs create AI enabled flows.
- Help NGOs understand and make use of the data from interactions with their chatbots
- Help NGOs enhance their existing chatbot programs
~32 participants, from across 19 orgs joined the sprint in Mahabalipuram. The spirit of the gathering was to invite NGOs and folks from the NGOs who are hands-on, and create the agenda to be able to curate their own learning experience and as much as possible. Utilize the 2 days of in person time to create something new, apart from making connections with the community that is using chatbots for social impact.
Below is the perspective of the Glific team capturing the essence of what their respective NGO contacts worked on.
Rocket Learning
Rocket Learning empowers 2.5 million children in India with quality early childhood education through digital communities using WhatsApp and YouTube, partnering with the government to scale affordable ECE solutions. Their approach builds parental and teacher engagement, aiming to impact 10 million children in the next five years.
Amit Badoni joined from the Rocket Learning team as a sole representative, and had two key goals: analyzing audio quality to prompt users to retry queries when needed and retrieving the correct educational link through voice queries. He tackled the first by leveraging Glific’s voice file search and analyzing audio quality using Bhashini’s speech-to-text, then scoring the text semantics with ChatGPT. By integrating both into a single seamless flow, he successfully demonstrated the solution, marking a step forward in improving voice-based learning interactions.
Sol’s Arc
Sol’s ARC provides inclusive education solutions to empower socially and economically disadvantaged groups, improving their learning and employment outcomes. They build an ecosystem that supports struggling learners across their lifespan to break the cycle of poverty.
Jerin joined the Mumbai team of Sol’s Arc and set out to detect potential learning disabilities through a child’s handwriting. Leveraging Glific’s GPT-Vision feature, he successfully made the flow that can identify early indicators of dyslexia, ADHD, and other learning challenges. He also figured out that for more accuracy he can try GPT vision fine-tuning later at scale.
Madhi Foundation
Madhi Foundation is an NGO focused on improving foundational literacy and numeracy in India’s public education system. They work with governments and schools to bridge learning gaps and ensure quality education for all children.
Gokul Kumar and Porvika joined from the Madhi team. They were able to create a chatbot conversation which can query using Tamil audio and give adequate results back in Tamil audio. They were able to do this using voice-filesearch.
Haqdarshak
Haqdarshak is an NGO that helps citizens access government welfare schemes and benefits by providing digital and assisted solutions. They empower communities, especially low-income and marginalized groups, by improving awareness and facilitating application processes for social security programs.
Lakhanveer Lodhi joined from the Haqdarshak team. From having deployed a schemes delivery chatbot that focused on text input and output, he focused on the learning that many contacts just had posters of the schemes they saw posted on railway stations, busstops etc. As a next iteration on their ongoing schemes chatbot, he was able to add functionality where contacts could take a photo of a govt scheme (which can be a roadside poster or online images) and the chatbot then provides details about that scheme. He was able to do this with a combination of gpt-vision and filesearch.
Piramal Foundation
Piramal Foundation works to improve the lives of marginalized communities in India by partnering with governments, organizations, and academia. It focuses on empowering youth and strengthening state systems to create impactful initiatives, benefiting over 11.3 crore people in 16 years.
Naveen Kumar and Ayush Dwivedi from Piramal Foundation joined in at the sprint. They built a prototype to provide feedback to students and share relevant video links The system uses a set of questions to assess students’ competencies, such as leadership and resource management, assigning scores accordingly. Then the scores are sent to OpenAI to generate feedback, determine the students’ levels based on the scores, and provide appropriate URLs for further learning.
Vipla Foundation
Vipla foundation focuses on early interventions and system strengthening in education, health, child protection, and disability sectors. They started using Chatbot from 2024 to share activities/ resources to Teachers and Parents on early child care.
Saee and Jaymala from their organization attended the Sprint. Although they joined the Sprint as part of the Data track to understand how to extract data from the chatbot for reporting, the event provided them with a broader experience. They had the opportunity to explore various features and possibilities within Glific, which enhanced their existing use case and opened up additional use cases for them. Additionally, they strengthened their foundational understanding of the platform, identified gaps in their current flow structure, and are eager to further enhance their implementation.
eVidyaloka Trust
eVidyaloka Trust provides quality digital education to rural children in India through online volunteer-led classes. It partners with local communities and government schools to bridge the learning gap using technology. Previously, they used a combination of WhatsApp groups, one-on-one chats, and calls to onboard volunteers, share materials, and provided class management tips. However, this approach requires significant effort and leads to scattered communication. To address this, they recently introduced a chatbot as a centralized platform for volunteers to engage with throughout their journey. They attended the Refresh Track to strengthen their foundational knowledge and build chatbot flows to facilitate a smooth launch upon their return.
Nikitha and Nitya are enthusiastic learners who embraced every opportunity during the Sprint to acquire new knowledge, reflect on their flows, and implement improvements. During the Sprint, they redesigned their entire onboarding flow for various stakeholders (volunteers, donors etc), incorporating best practices and defining clear next steps for their implementation.
Sauramandala
Sauramandala addresses the challenges faced by remote and inaccessible communities by accelerating social and economical change through collaborative action. We cater to remote and inaccessible communities with specific solutions pertaining to the needs of the people.
Gladinia and Ritika Chetri joined from all the way from Shillong. Apart from a minor challenge in setting up and configuring GCS, they developed a prototype that successfully read data from a Google Sheet. By incorporating a counter in their workflow, they were able to deliver content continuously. They recognized the need for this functionality to track user progress and support the development of a dashboard. They also hosted a quick 10 mins session to share about their work of preserving the narratives and culture from the north east, and distributed with the rest of the NGOs lovely souvenirs like story books and stickers.
INREM
INREM Foundation empowers the rural communities of India by building Water-Safe Communities. We achieve this through policy support, access to knowledge, and working out innovative solutions to help them overcome the challenges of contaminated water bodies.
Kiran from INREM joined at the sprint as a part of the Data Track. After re-formulating a hypothesis to crowdsource water quality data from a new target group – front line workers, he built a simplified registration flow and data collection flow. The next steps of this involves visualizing the data being collected through these flows.
Reap Benefit
Reap Benefit works is igniting a movement of change makers to redefine civic and climate leadership one action at a time. Athira T J, Data Analyst and a first time Glific sprinter as well as and Gauthamraj Elango, the Head of Technology at Reap Benefit, on his sophomore sprint joined in for 2 days of making and sharing.
Athira started with the aim to create an AI mentor that would be armed with knowledge from Solve Ninja Forum. Reap Benefit’s open source knowledge repository for citizen action. This AI mentor would direct/ nudge young people to take next steps in their problem solving journeys based on the design thinking, and behavioural science framework by Reap Benefit through years of iterations and implementing solutions for real world problems.
SNEHA
SNEHA (Society for Nutrition Education and Health Action) holistically addresses preventive and promotive healthcare with special focus on vulnerable women and children. Priya Verma, Product Manager and Nilotpal, CTO of SNEHA joined in at the sprint from Mumbai. For both of them this was the first Glific sprint.
Priya Verma being a recent addition to the SNEHA team came in with the aim of getting in the weeds of AI flows so that she is able to meet the emerging and growing needs of SNEHA’s explorations of AI programs on ground. She prototyped a flow that can be used to test on golden questions (a collection of known questions and answers). This would help the SNEHA team to get a feel for how the responses are being generated by the AI Assistant for different iterations of a prompt or upon changing knowledge base. As a beginner on Glific (less than 2 months of hands on using) Priya was able to understand and implement the use of google sheets to read questions, write to google sheets, create new ai assistants and implement iterations using contact variables.
Antarang Foundation
Antarang envisions a world where all young adults are passionately, positively and productively engaged in careers of their choice. Cibel Mascarenhas and Gayathri RS joined in from the Mumbai team of Antarang. Both first timers at the sprint were looking to leverage this time to create a solution using AI for the upcoming career, new academic year starting season.
They observed through their experience that there is a high need for career guidance as the school year draws to a close. Students seeking information around which pathway to choose post 10th, post 12th, post diploma, in order to make progress towards a particular career. And there exists a lack of enough knowledge around them. Cibel and Gayathri put their heads together and were able to create an AI assistant that incorporated knowledge of 51 career pathways. The first prototype was tried by the NGO peers during the showcase time, and by their own admission helped Gayathri and Cibel to test interaction experience and learn about the users expectations.
Key Education Foundation
Key Education Foundation builds the capacity of government systems to deliver quality Early Childhood Education. Namratha Bhat and Swarupa from the Product Development team of KEF joined in from Bangalore. Being a long time user of the chatbots program delivery, KEF team were aiming to reduce the time it takes for a teacher to complete the milestones of delivering their CLAP program in respective schools.
As a way of monitoring the program delivery, teachers across hundreds plus schools are required to keep a record of the dates, no of worksheet shared to the parents on a weekly basis and so on. Teachers are required to submit this data through forms. KEF team ideated and implemented the use of Vision AI to reduce the repetitive work from teachers. Through multiple iterations, KEF was able to come up with a template and a prompt that reliably extracts the handwritten information by the teachers and pulls it into a google sheet.
India Literacy Project
India Literacy Project is dedicated to enable children in government schools to attain grade appropriate skills and create pathways to productive careers. Hemalatha Ambalavanan and Swethamalaya.K from ILP Chennai participated in the Glific Sprint Over the two days of the Glific sprint, they were able to achieve the following outputs:
1. Integrated Google Sheets to send weekly messages to Class 12 students in their collections via triggers.
2. Leveraged AI to provide students with answers to their questions.
3. Developed a new Career Counsellor workflow for village-level volunteers, whom they can train to spread career information to students from Class 9 to 12.
4. Implemented a registered split flow in the main menu so that registered students no longer need to re-enter their details.
Waste Warriors
Waste Warriors is dedicated to sustainable waste management and environmental conservation in India through community engagement, education, and waste collection initiatives. They introduce the WhatsApp-integrated EcoDaan Chatbot, designed to inspire tourists to take small yet impactful environmental actions. By simply scanning a QR code, travelers can join the EcoDaan movement and become part of an experiential eco-community initiative.
Anjali Mehra and Gaurav Chetal from the Waste Warriors team joined in the sprint. They used the split function for the first message into three different flows with a different and tailor made first message nudge for those flows. They also understood the importance of `optin` message and saving it in the starting of the flow and accordingly tweaked their flow. Additionally they also closed all the dead ends. After the LLM session they were determined to design a flow based on LLM models as well.
FMCH India
Foundation for Mother and Child Health, works with the vision to end preventable child malnutrition in india. Ankita Boral, a nutritionist by education but representing the Products team from FMCH was Ankita. Through the sprint she found herself divubg deep into the world of SQL, asking the right questions like, is the chatbot able to help FMCH identify malnourished children? and imagining what the visualisation of such data should be like for decisions at various levels of the program. Being the sole representative from FMCH, through the sprint Ankita could be seen engaged in intense conversations with her team over meet calls to define what does engagement mean for FMCH, and dwelling juggling between Big Query tables and Glific flows to pinpoint what the FMCH team wants to see when they say retention rates. We hope this helps the FMCH team move one step closer towards building a stronger data culture.
Lend A Hand India
Lend A Hand India, aims to integrate vocational education with mainstream education at secondary and higher secondary education levels (Grades 9-12 and student Age 14 to 18). Ravindra and Jayesh were present from the LAHi team. Both were first timers at the sprint. Ravindra could be seen prototyping a chatbot that enabled the vocational trainers, how knowledge consumption and contribution through the chatbot can help them who would have limited time for training, need to handle subjects outside their expertise, and face high workloads while managing multiple subjects, to access the training material that they need to execute training sessions in schools, as well as contribute to the repository of knowledge on vocational training being maintained by the LAHI team.
Their prototype could potentially help grow the knowledge repository on vocational training from on-ground practitioners. The chatbot was enhanced to evaluate whether shared content (via voice or text) is relevant/irrelevant and when requested for information it delivered accurate information in most cases. However, when the assistant attempted to provide answers beyond the available resources, the prompt was fine-tuned to ensure responses remained strictly within the provided content.
Leadership for Equity
Leadership for Equity is helping governments deliver a bright future for all children. Anirudh Prasad, Director at LFE joined from the Pune team. Anirudh has been liasoning the conversations between Glific and LFE over the course of last few months and he came in with the purpose to get more hands on and get a more nuanced understanding of the present workflows. Anirudh was seen enhancing the flows for onboarding and delivery of AI assistant driven content support for teachers depending on the geography the teacher belongs to. This required a slightly complex mapping of serving the right course options to the teachers based on their geography and then ensuring that an AI assistant with the right knowledge base is evoked to answer the teacher’s inquiries.
Anirudh also ideated the plan for the next 3 months of roll out of chatbot to the teachers. These plans included onboarding ~10k teachers and then sending notifications that increase the total number of course completions that the teachers themselves are enrolled into.
Make A Difference
Make A Difference (MAD) provides holistic care and customized programs for children in need of care and protection. They aim to ensure equitable outcomes for these children, comparable to middle-class standards. Their focus is on creating a life free from exploitation and building resilience to cope with life crises.
Chetan built a prototype where volunteers answer weekly questions, and their responses are sent to OpenAI. The system already has the questions and correct answers stored. Based on the volunteers’ answers, OpenAI provides feedback on areas for improvement. Each week’s data is stored in a contact variable in Glific to maintain a record of previous feedback. This allows OpenAI to consider past performance when giving future feedback by sending the contact variable data to openAI.
Common Sessions
With a sprawling agenda, spread across 3 themes: AI, Data and Refresh (enhancing of chatbot) following sessions were conducted as a way of knowledge sharing:
- Influencing user behaviours
- Glific best practices (checklist)
- Basics of LLMs in Glific
- Prompt Engineering
- Introduction to BigQ Tables of Glific
- Right Question To Ask (building effective dashboards)
- Talk by Reap Benefit on using Technology as an Enabler
- ChatMnE Talk by 10x Impact
Conclusion
As the NGOs using Glific increase over a period of time, it will be crucial for Glific to find a way to get more in touch with the diversity of people using the platform. Diversity of language, geographies, causes and most importantly of all the level of tech-friendliness. The sprints serve as the best reminder of the problem statements tech has to address, and the reason for its existence.
