.svg)
What happens when you bring a room full of people building with AI together, give them a stage, and ask them to show what they’ve been working on?
You get ideas that move beyond conversations. Live demonstrations. Unexpected discoveries. Plenty of questions. And, occasionally, a few AI-generated Halloween memes.
That was the spirit of SWARM 2: Show and Tell, the second edition of SWARM AI community, Powered by Galent, held on Friday, October 30, at Galent’s Chennai office, with the participants also joining in virtually.
SWARM is built around a simple belief: the best way to understand what AI can do is to see what people are doing with it. And this edition put that belief front and centre. Instead of a lineup of presentations about AI trends, the spotlight was on people building real things with AI and sharing their work, experiments and learnings with others.
Before the demos …
The evening began with a challenge that had nothing to do with writing prompts or debugging code.
Every attendee received a networking bingo card. Each square came with a little mission: get up, meet someone new, find out what they’re working on, and tick off the challenge. The first five people to complete their cards would walk away with prizes.
It was a fun way to break the ice, especially in a room where not everyone knew each other. Because while SWARM is about building with AI, it is just as much about building connections between the people doing it.
And since it was Halloween weekend, the community had another challenge waiting.
#SwarmFrightNight
A QR code invited attendees to join the SWARM WhatsApp community and take part in a Halloween-themed meme challenge. The brief was simple: create a meme on your phone, post it with #SwarmFrightNight, and let the community decide the winner through reactions.
The prompts were wonderfully familiar to anyone who has spent time building with AI:
- “When the AI hallucinates with full confidence.”
- “Me at 4:55 PM asking AI to fix a bug.”
- “Ghosting: when the chatbot leaves you on read.”
It was a reminder that an AI community doesn't have to be all serious conversations and technical deep dives.
Setting the tone
With the room warmed up, Sriram Rajagopal, Founder of Galent, opened the session with a short address on AI and the importance of creating spaces where people can explore, experiment and exchange ideas.
The spirit of the evening was clear: innovation doesn't happen only in formal labs or planned projects. It happens when people try something, learn from it, challenge an assumption, and share what they discover.
And then came the part everyone had been waiting for: the show and tell.
Three builders. Three different takes on building with AI.
The heart of SWARM 2 was its contributors - Galentians who stepped up to demonstrate what they had been working on. Each brought a different perspective on applying AI to real engineering challenges.
1. Rethinking the AI testing experience
Mohanram | Senior Artificial Intelligence Engineer, Galent
What if a Jira ticket could set an entire AI-powered testing team in motion? Mohanram’s Agentic AI Testing Marketplace brings together 21 specialist AI agents to handle functional, automation, performance and security testing, coordinated by an orchestrator with quality checks at every stage. In one demonstration, the system processed seven acceptance criteria and saved 14 test cases to Jira in just 283.8 seconds.
Curious how the AI testing team works?
Click here for the full walkthrough →
2. Exploring Superhuman Mode
Siddarth S | Infrastructure Automation Engineer, Galent
Managing 655 Windows servers comes with a mountain of repetitive checks. Siddarth’s Superhuman Mode uses AI-assisted automation to streamline patching workflows, cutting down manual effort across pre-checks, post-checks and reporting. The scale of the opportunity? Around 218 hours of pre-check effort every month. With eight of 12 automation tasks already running, the journey is well underway.
Want to see how Siddarth put AI to work?
3. Making quality engineering more autonomous
Ramkumar Tulasidharan | Senior QA Architect, Galent
Can AI turn months of test-design work into hours? Ramkumar’s Autonomous QE workflow brings application understanding, dependency mapping, test generation and automation into one connected process. In a demonstration involving 88 screens, around 200 dependencies and 469 test cases, the team showcased how an estimated six months of manual test-design effort could be reduced to approximately six hours.
Discover how autonomous quality engineering is taking shape.
Prompt Roulette
Just when the technical conversations had everyone thinking, the evening took a playful turn with Prompt Roulette, the Halloween edition.
The idea was to put prompting skills and a little creativityto the test. Participants took on the challenge of working with prompts and seeing what AI could produce. It was a lighter moment, but it also brought the community back to one of the most interesting parts of building with AI: the possibilities can change dramatically depending on how you approach a problem.
Communities make the difference
To close the evening, Murali Gunashekaran, Vice President, AI & Emerging Technology at Galent, spoke about the importance of communities in the fast-moving world of AI.
Technology may be advancing at an extraordinary pace, but learning doesn't have to happen in isolation. Communities create the space to exchange knowledge, discover new approaches, learn from each other's experiments and find people who are asking similar questions.
And then it was time for coffee, conversations, a few more connections and the announcement of the networking bingo winners.
SWARM 2 was a reminder that building with AI is not just about having access to the latest tools. It's about curiosity, experimentation, sharing what works, being open about what doesn't, and learning together.
Have something you've been building with AI? We'd love to see it.
Join the SWARM community, meet fellow builders, and put your name forward to demo at the next edition.
JOIN THE SWARM COMMUNITY →
.webp)
.webp)
