Who Is Kari Briski at Nvidia?

Kari Briski is NVIDIA's vice president of generative artificial intelligence (AI) software for enterprise. She leads the Nemotron open model family and the software around it, and she joined Nvidia from IBM in 2016. In October 2026 she is the executive telling developers that Nvidia will keep releasing newer and better open models.

Who Kari Briski Is

Kari Briski is vice president of generative AI software for enterprise at Nvidia, according to her Nvidia author page.

Briski leads the Nemotron AI model family and the software ecosystem around it.

On October 15, 2026, Briski bylined Nvidia's announcement of Nemotron 3.5 Super VL and Nemotron 3.5 Bolt. Two days later, Digital Today reported her saying Nvidia "will continue to release newer and better open models."

• Current title: Vice President of Generative AI Software for Enterprise at Nvidia
• Company: Nvidia (joined in 2016)
• Leads: the Nemotron open model family and the software around it, plus strategy and roadmaps for Nvidia's AI software and services
• Previous employer: IBM
• Education: Bachelor's degree in computer engineering from the University of Pittsburgh
• In the news for: Launching Nemotron 3.5 Super VL and committing to ongoing releases of open models

Executive Role and Product Portfolio

Briski oversees strategy, product roadmaps, and requirements for Nvidia's artificial intelligence software and services.

Her current official title on the Nvidia author page is Vice President of Generative AI Software for Enterprise at Nvidia. External event listings and trade coverage show minor variations in this designation. MIT Technology Review's EmTech speaker page listed her title as VP, AI Models, Software, and Services under the name Kari Ann Briski, while the Cohesity Catalyst speaker directory listed her as VP of Gen AI. Media publications such as Computerworld in July 2026 designated her as vice president of generative AI software, and Digital Today in October 2026 named her vice president of enterprise generative AI software.

Within this executive remit, technology outlet Digit reported that Briski leads the Nemotron AI model family and the software ecosystem built around it. In March 2024, The Register reported that she directed software product management for foundation models, libraries, software development kits (SDKs), and microservices for training and inference, including NVIDIA Inference Microservices (NIM) and NeMo.

Briski has published 18 articles on the Nvidia blog, according to her author page. Her 2026 posts cover Nemotron 3.5 Lightning and NeMo Switchyard (August), Nemotron 3 Nano Omni (April) and Nemotron 3 Super (March). In May 2026 she wrote about Nvidia and ServiceNow agents. Her earlier posts covered new open models and datasets in January 2026, a partnership with Mistral AI in December 2025, and the UK-LLM initiative in September 2025.

Career History of Kari Briski

Briski built a multi-decade career in software engineering and product management before taking on enterprise generative AI at Nvidia.

She holds a bachelor's degree in computer engineering from the University of Pittsburgh.

Briski worked at IBM prior to joining Nvidia in 2016. Her official biography on the Nvidia blog states that she brings more than 25 years of software development experience and 20 years in technology product management. While an older speaker biography for EmTech listed 20-plus years in software and 15 in product management, the figures published on Nvidia's website represent the current record.

In March 2024, The Register detailed her work directing product management for artificial intelligence and high-performance computing (HPC) SDKs. She told the outlet that NIM lets developers "get from zero to a large language model endpoint in under five minutes." On internal AI use, she said "there were 150 teams within Nvidia that I found."

Enterprise Commitment to Open Models

Nvidia reinforced its public commitment to shipping open foundation models with the October 15, 2026 launch of Nemotron 3.5 Super VL and Nemotron 3.5 Bolt.

On October 15, 2026, Briski authored the official announcement on the Nvidia blog detailing the launch of Nemotron 3.5 Super VL and Nemotron 3.5 Bolt. In that announcement, Briski noted that Super VL was built in collaboration with the Nemotron Coalition. Technical capabilities of that release are covered in our guide to Nemotron 3.5 Super VL explained.

Following the release, Digital Today reported on October 17, 2026, that Briski affirmed Nvidia will continue to release newer and better open models. In indirect speech, the publication noted her statement that public agencies and financial institutions are reluctant to use China-made models.

On October 17, 2026, JKN reported, citing an October 16 report in The Wall Street Journal (WSJ), that Briski said: "We want to send a message to developers, partners, and corporate customers." JKN quotes her adding that "they can continue to trust and utilize Nvidia."

Strategic Rationale and the Software Analogy

Briski explains Nvidia's open-model investments by comparing model weights to core computing libraries like CUDA.

Speaking with Digit on October 18, 2025, Briski explained: "We believe that to further AI development you have to have open models." In the same interview, she noted that "Nemotron is not new, I think it's NVIDIA's best-kept secret," while observing that "There's no one model to rule them all." Comparing the model family to Compute Unified Device Architecture (CUDA), she stated: "this is like CUDA, like our math libraries, you can depend on it."

In December 2025, Reuters reported Briski's remark that Nvidia aimed to provide a "model that people can depend on," explaining: "This is why we're treating it like a library." Speaking to Gizmodo that same month, Briski added: "When we're the best development platform, again, people are going to choose us."

Briski expanded on this commercial logic in an April 2026 interview with The Deep View, stating: "The model is the byproduct." She noted: "You need a big open ecosystem for everybody to come along." In an article on the Nvidia blog on March 25, 2026, Briski wrote that "Proprietary versus open is not a thing" and observed that "Openness fuels innovation across the model ecosystem." Later in July 2026, she told Computerworld that "Open models allow enterprises and countries to own, inspect, and adapt models with their own data." She explained: "When you put a model out into the open, you get more startups, more builders, lower entry barriers," and clarified: "We're building Nemotron first and foremost for ourselves."

Interviews, Podcasts and Speaker Listings

Briski has discussed Nvidia's models on a developer podcast, in a 2024 interview with The Register at Nvidia GTC, and on two event speaker lists.

On March 10, 2026, Briski was the featured guest on the Stack Overflow Podcast to discuss the Nemotron model architecture, hardware-software co-design, and low-precision model training workflows.

The EmTech event page lists her as a speaker on "Using Generative AI to Tackle Global Challenges." Cohesity Catalyst also lists her as a speaker.

BornCity reported on July 28, 2026, in indirect speech, that Briski said only open models allow the inspection needed for effective defense. The report covered the Nvidia-led Open Secure AI Alliance.

Strategic Considerations for Enterprise Model Selection

Briski's CUDA comparison and her October 2026 statement that Nvidia "will continue to release newer and better open models" tell a buyer that Nvidia plans to keep shipping new Nemotron models. What each Nemotron release changed is covered in our guides to Nemotron 3 explained, Nemotron 3 Super explained, and Nemotron 3 Ultra explained.

At Layer3Labs, we build custom agents and enterprise integrations for business workflows, and the choice between open models and closed APIs centers on control. When an enterprise must audit model weights for regulatory compliance or run inference entirely within private environments, open weights provide direct inspection rights.