Sponsored by Marta DeBellis, CMO of Ataccama and founder of the AI Council initiative, the hackathon brought together internal coaches and Microsoft experts to help Ataccamers push boundaries and turn bold ideas into action.
Behind every successful hackathon is a dedicated organizing team. This initiative was led by Corey Keyser, Head of AI at Ataccama, with support from local coaches Daniel Ryan, Marek Ovcacek, and Daniel Marinovic.
“Our teams had so many ideas on functional use cases where they wanted to build AI Assistants or AI Agents to accelerate and improve these tasks but they struggled to find time in their day-to-day work. We started last year with a Digital SKills Academy teaching employees how to use Generative AI, but found they needed more support and a dedicated sprint set of days to make progress.
The Marketing Team had 6 teams in the hackathon and built an Agent to build brand images, a demand generation KPI optimization tool and an Agent to build hundreds of account-based custom landing pages.” says Marta DeBellins, CMO of Ataccama.
Prior to the hackathon itself Ataccamers were offered the opportunity to consult on their projects, participate in workshops and trainings and prepare themselves for hacking! I asked Corey a set of questions to learn insides from behind the scene.
What is the role of the AI Council at Ataccama, and how did it contribute to the organization of this hackathon?
The AI Council is responsible for two main things: encouraging effective use of AI across Ataccama, and ensuring we have solid governance in place for responsible AI practices. As a member of the AI Council I organized the hackathon as a way to introduce AI in the company. We saw it as a great way to introduce AI to teams outside of product and engineering—especially through low-code and no-code tools—so they could start solving their own day-to-day challenges with AI.
What was the main purpose behind launching this AI Hackathon? Was it more about innovation, team building, or uncovering new product ideas?
Honestly, it was a bit of all three, but our biggest driver was education and accessibility. We wanted to help non-technical teams get hands-on experience with AI tools, and also give them enough context to understand what AI is actually good at. That said, it also turned into a fantastic team-building opportunity. I saw a lot of excitement in the Toronto office—people got together, collaborated, shared meals, and just had fun. It created a lot of cross-functional energy that you don’t always get in a typical workday.Can you share examples of some standout projects or problem statements teams tackled during the hackathon?
Every team did something interesting, but a few stood out:
- First place went to Team Built Different — John Efechaobor, Saaraa Alli, Rafeeq Olawale Durowoju, Anna Kitor, MEng, and Abir Newaz — for building Client 360, an AI Agent that integrates five data sources using Graph RAG. It enables CSMs, Sales, and Consultants to search and summarize customer information quickly from a single source, eliminating the need to sift through Slack or message colleagues. A powerful boost in efficiency for customer-facing teams.
- Second place was claimed by Team Imagery Imagined — Sandeep Daley, Maria Echeverria Maldonado, Kat Dominiak, Hyunjoo Park, and Martina Sulkova — with Ascent, an image generation tool integrated with Asana. It allows Marketing and Design team members to generate brand-aligned visuals on demand, streamlining the creative process and incorporating approval flows directly into existing workflows.
- Third place went to Team Ken — Nathaniel Reid-Smith, Kivanc Aykac, and Emad Raza — for developing a smart lead qualification tool. Using a multi-agent architecture, their solution powers a chat assistant that helps Sales quickly assess and prioritize new leads based on internal databases and value stories—an invaluable tool for pre-sales efforts.
We had fantastic projects across all our departments with a lot of participation from People & Culture, Sales, Customer Support, Services, and even Engineering. For example our people and culture team created agents for answering HR questions and supporting career development. And our CEO Mike created a custom GPT to help sales teams find relevant customer value stories—so cool to see him jumping in!
Who were the judges, and what were the main judging criteria used to evaluate the projects?
The judges were the four coaches: myself, Daniel Ryan, Marek Ovcacek, and Daniel Marinovic. We scored each project on:
- Value of the solution
- Clarity of the problem they were solving
- Presentation quality
- Inventiveness—how creatively they approached their technical solution
What was the biggest challenge in organizing or running the hackathon—logistics, engagement, scope of ideas?
The biggest challenge was definitely managing the wide range of technical abilities. Some teams had no coding experience at all, while others were highly technical. We had to provide resources and support for both ends of the spectrum. But I think we struck a good balance—there was something for everyone.
What are the key takeaways from this experience that you’d share with someone organizing a similar event at another company?
My biggest takeaway is that anyone can build an AI tool that helps them work smarter. You don’t need to be a developer. With the right tools, training, and encouragement, anyone can create something valuable. That's the mindset we want to foster—AI-first thinking, across all teams.
We're planning to keep this momentum going. We'll probably continue with a technical hackathon in Prague in the fall, and then a more cross-functional one like this again in the spring. It’s a rhythm that really works for us.
What's one recommendation you'd give to teams participating in future hackathons at Ataccama?
Don't underestimate what you can do. Even if you’ve never touched a line of code, you can still build something great. The key is to start with a real problem you care about and then just experimentAlso, some of the image model projects were more technically challenging than expected, so I give a lot of credit to those teams for sticking with it and delivering great results.