At NVIDIA, ChatGPT Work is helping turn repetitive tasks and information overload into time for making decisions. The company says its sales, marketing, and solutions architecture teams are already using it to organize processes, analyze AI developments, and create prototypes more quickly.
Less Manual Work to Prepare for GTC
Will Daney works with NVIDIA’s global sales, business development, and product teams. One of his responsibilities is supporting the planning of GTC, the company’s global artificial intelligence conference.
In the past, preparing for the event involved gathering account lists, reviewing records, and helping teams decide what actions to take with customers and partners. Much of that analysis was done manually in spreadsheets and could consume nearly 40% of his time.
Now, Will has turned much of the process into an automated workflow with ChatGPT Work that runs twice a week. During the 12 weeks of GTC planning, the system saves approximately 16 hours of work each week.
The benefit isn’t just completing a task faster. It also allows Will to spend more time talking with field teams and helping them improve the customer experience.
One important advantage is that Will can adjust the workflow himself when the event’s needs change. He doesn’t have to wait for a new tool to be purchased, implemented, or maintained.
The process can also be shared with other regions. Teams in San Jose, Taipei, Europe, and Washington, D.C., received his workflows and adapted them to their own operations. It’s an example of how one person’s knowledge can become a reusable process for an entire organization.
From Reading Updates to Finding Useful Signals
Rachita Jain is part of the artificial intelligence operations team within NVIDIA’s marketing organization. Her challenge is keeping up with an industry where new models, benchmarks, and research appear practically every day.
The problem isn’t finding information. The real challenge is knowing which developments matter to NVIDIA and how they relate to its projects, conversations, and internal priorities. What’s the point of reading dozens of news stories if none of them leads to a concrete decision?
To solve this, Rachita created a workflow with ChatGPT Work that reviews reliable external sources alongside the company’s internal information. The system identifies important connections and summarizes between 25 and 40 AI updates each week into 5 to 8 actionable signals.
The idea is to move from passive reading to active intelligence: not just knowing what happened, but understanding why it matters and what could be done about it.
Prototypes in Days, Not Weeks
The same environment also helps develop new ideas. Rachita can explore a concept, evaluate different approaches, work with code, fix errors, and improve the result without constantly jumping between disconnected tools.
In one case, she went from an initial idea to a working prototype in approximately 3 to 5 days. Previously, she estimated that manually building the components and connecting them across different applications would have taken between 2 and 3 weeks.
This doesn’t mean AI eliminates technical work. It means it reduces the friction between stages: researching, programming, testing, and fixing. When those tasks are better connected, an initiative that might once have remained an experiment has a better chance of becoming a functional product.
Knowledge Becomes Reusable
NVIDIA’s goal goes beyond saving individual employees hours. The company wants to turn specialized knowledge into workflows that other teams can adapt for different functions, events, and regions.
This approach keeps people close to the process. Employees who understand the operation can modify the instructions, adjust the criteria, and decide how the automation evolves. AI works as a support layer, not as a black box that no one can change.
The goal is for teams to spend less time gathering information and more time interpreting it, collaborating, and helping customers.
NVIDIA’s example shows a practical way to bring AI into a large company. You don’t have to start with a complete transformation. Sometimes, the first step is identifying a task that repeats every week, turning it into a clear workflow, and measuring how much time the team can recover.
When those processes are shared and adapted, one person’s experience is no longer isolated. It can become a tool that helps more employees work with less operational burden and focus on higher-value decisions.