Building AI from the Silicon Up: What It Takes to Train and Run Large Models at Scale
Most AI conversations stay at the surface. This one starts at the chip level and works its way up to production.
We are sitting down with Natalia Vassilieva, VP and Field CTO at Cerebras Systems, the company behind some of the fastest AI infrastructure in the world. From Professor at St. Petersburg State University to VP and Field CTO at Cerebras, Natalia has spent her career at the intersection of algorithms, hardware, and what it actually takes to run AI at scale. She knows what breaks, what scales, and what the industry rarely talks about publicly.
This is not a passive webinar. We want cameras on, mics open, and real questions from the room. You will be joining a community of women from all over the world building tech companies in their countries. Come ready to connect, not just listen.
Here's what we're getting into:
Why hardware design choices make or break AI performance and what led Cerebras to rethink the chip entirely
The thinking behind open sourcing Cerebras-GPT and what the team learned from it
What the path from research idea to production AI system actually looks like
The conversation will be moderated by Technovation alumna Soumya Tejam, who is on the GTM team at Sardine AI. She previously backed underrepresented founders at Harlem Capital and climate-tech startups at Blackhorn Ventures, and is passionate about creating pathways into tech for young women innovators.
Format
We will begin with a moderated Q&A for 20 minutes and move to an open discussion. Come with questions!
Register to get the access link one hour before the session.
If you are a Technovation alumna or Women Techmaker Ambassador and cannot register, please email [email protected].