For a change, it was hot in Bangalore this time, even though peak summer hadn’t started. This was my third time flying to Bangalore for a Glific event, but this one felt different.
It was my first Accelerator event, with a cohort of 34 NGOs.
This time, the focus was clear: integrating AI into chatbot programs.
These 34 organizations, working across different domains, geographies, and beneficiary groups – came together with one shared goal: to design, build, pilot, and launch AI-powered WhatsApp chatbots for real-world impact over a period of 6 months.
Want to dive deeper into the cohort? Refer to this guide to explore the participating organizations and the AI chatbot use cases they are building.
Day 0: Welcoming, Icebreakers & Inspiration
We, the Glific team, arrived a little early to get everything in place. From setting up the venue to aligning on the plan, it was all about preparing the space, and ourselves, for the accelerator.
Right after a quick team huddle, we did what we always do – yelled “Accelerator!” at the top of our lungs.
A small ritual, but it instantly boosted our energy (and set the tone for the next two days :p)
As participants started arriving at the venue – Fireflies, tucked away from the city chaos and surrounded by greenery, it already felt like a space meant for meaningful conversations (though the daytime heat had other plans!).
With around 80 people (including the Glific team), it was important to build connections before diving into work.
The event began with Radhika’s welcome address, followed by a fun and engaging icebreaker activity. Participants were grouped based on their answers to questions, and these groups kept rotating, ensuring everyone got a chance to meet each other.
The day ended on an inspiring note with Shruthi from ATREE, who showcased:
- How they are using AI for ecological practices
- Their approach to evaluations using Kaapi
- Their vision for the future
Day 1: Fresh Energy and AI in Action
The day started with coffee, conversations, and a refreshing morning breeze. The atmosphere was full of energy, participants were actively networking and exchanging ideas even before sessions began.
Mentoring Across Diverse Impact Areas
Out of the 34 NGOs (10 existing and 24 new), each Glific team member was assigned mentoring responsibilities.
I had the opportunity to mentor three NGOs:
- Fadhila Health – Focused on reproductive, maternal, neonatal, and child health (RMNCH) in Kenya
- ResGov – Focused on governance and public service delivery
- Peepul – Working to improve learning outcomes in India’s government schools by empowering teachers
These represented just a fraction of the diversity, across health, governance, and education, seen in the entire cohort.
Building AI Assistants (Hands-on!)
The first hands-on session of the day focused on creating OpenAI Assistants within Glific. The examples used made it immediately engaging, and the room was buzzing with questions from peers.
It wasn’t just a session to watch, everyone got hands-on with assistant creation, from writing prompts and importing knowledge base files to integrating everything within a flow.
What made it impactful:
- Real world examples
- Interactive Q&A
- Immediate hands-on practice
Before participants jumped into their own use cases, Udhyam, an education-focused NGO shared how they are already using Glific and leveraging AI to reach and support their beneficiaries. It sets the perfect context.
Turning Ideas into Working Chatbots
With the fundamentals in place, every organization was eager to make it work for their specific program. Each NGO came with their knowledge base ready, and a bot set up by the Glific team helped them generate and refine prompts according to their needs.
With mentoring support, they:
- Generated and refined prompts
- Structured their knowledge bases
- Built chatbot flows
- Integrated AI assistants
By the end of Day 1, most NGOs had:
- AI assistant created
- KB and prompts refined
- Flow designed
- Chatbot tested on phone
The phone testing was the highlight of the day. Seeing real responses in a chatbot built for their own beneficiaries brought a visible joy to the room. Everyone left the hall already thinking about how to make it even better the next day.
Day 2: Going Deeper with Voice & Refinement
The second day started warm, literally and figuratively. Conversations over breakfast were richer, more focused, and filled with ideas. (Also, I switched to chai, coffee isn’t really my thing!)
Exploring Voice: STT & TTS
Day 2 moved into more advanced territory, Speech-to-Text (STT) and Text-to-Speech (TTS). This was especially relevant because many NGOs’ beneficiaries prefer voice interactions.
The session, led by Tanu, walked participants through enabling voice queries and voice responses in their chatbots. It opened up a whole new dimension for the NGOs, and the room was immediately abuzz with ideas on how to implement it for their specific audiences.
A combination of rule-based flows and AI-powered flows started taking shape for many of the organizations.
Open Work Time: Build, Iterate, Improve
Open work sessions allowed NGOs to:
- Refine their chatbot flows
- Combine rule-based and AI-driven approaches
- Prepare for pilot launches
Mentors worked closely with teams to shape solutions aligned with their programs.
Showcase: Learning from Each Other
After two days of building, it was time to show. One of the most engaging parts of Day 2 was the showcase session.
NGOs were grouped into clusters of 4 – 5 organizations, and each group showcased their chatbot, talked about their program, and explained how the solution would impact their beneficiaries. The tables rotated so everyone had a chance to see what others had built, and pick up ideas to bring back to their own chatbot.
The tea break at the end of the day barely felt like a break. Conversations spilled over, How did you implement this? What’s your pilot plan? Who are your beneficiaries? and it was clear that two days had been more than enough to build a genuine sense of community among people who share a deep commitment to creating change. In just two days, a strong sense of community and shared purpose emerged.
What Participants Felt
Beyond the sessions, builds, and showcases, what truly stood out was how participants experienced the accelerator.
From hands on learning to meaningful collaboration, these reflections capture the essence of those two days:
“The knowledge I gained on chatbots and the support from the team were truly impressive.”
- Rama Mahato, Prison Aid and Action Research (PAAR)
“The workshop made me feel very comfortable with the platform, especially through the hands-on work time.”
- Srivathsan Ramaswamy, Nilay Foundation for Development Initiatives
“The collaboration and discussions with both the team and other NGOs helped us think through solutions effectively.”
- Akhil Shridhar, IBD India
Closing Thoughts: From Fireflies to Future Impact
As we said goodbye to Fireflies, minds were already on pilot plans and the impact ahead.
Even in the late evening, the sun didn’t seem to want to let go, almost as if the environment itself was encouraging us to keep going.
What stood out the most was not just the technology, but the collective intent: to use AI meaningfully, responsibly, and for real impact.
The next few months will be about turning these prototypes into real-world solutions, and I’m excited to see the change this cohort will create.
What we built in these two days is just the starting point. From here, organisations will begin by testing their chatbots internally and then running pilots with real users, learning from every interaction and refining their flows, prompts, and knowledge bases along the way.
As they move forward, the focus will shift to measuring performance through evaluations, improving based on feedback, and running larger pilots, ensuring that what they build is not just functional, but actually effective in real-world contexts.
By the end of this journey, these chatbots will evolve into more reliable, scalable systems, backed by learnings, guardrails, and clear impact stories, ready to be integrated into their programs and create meaningful change at scale.
Leave a Reply