Our Senior AI and Insights Engineer Zamena Jaffer has arrived in San Jose for NVIDIA GTC 2026, the world’s largest AI conference. Getting there wasn’t straightforward. Delayed flights, misplaced luggage, and more than a few anxious moments. But she made it, and last night she presented our research poster “Accelerated Speed to Insight” at the GTC poster reception.
The poster was selected by NVIDIA through a competitive review process, placing Adoreboard’s work alongside AI research from some of the biggest companies and institutions in the world. For a 13-person company from Belfast, that’s something we’re incredibly proud of.
The research covers a problem that sounds simple but isn’t:
How do you make sense of thousands of customer comments quickly, accurately, and at a scale that actually helps people make decisions?
Using NVIDIA A10G GPU acceleration, our approach achieves a 923x speed-up compared with manual human analysis. Processing time per customer comment drops from 30 seconds to 0.03 seconds. Overall analysis time is down 83%. And this isn’t a lab experiment. It’s in production, being used by enterprises right now.
This technology is the foundation of something we’re announcing this week: Customer Revenue Impact, a new way of connecting what customers say directly to financial outcomes.
Watch this space.
Zamena will be at the conference all week, attending talks, exploring new ideas, and talking to anyone curious about emotion AI, customer experience, or what happens when you point GPU acceleration at unstructured feedback. If you’re at GTC, come find her at the poster gallery.
Back home, the rest of the team is watching with pride and only a little jealousy.
We’re already making the case for a bigger delegation next year 😉
