Cresta: From Stanford Research to AI at Work
Congratulations to the Cresta team on joining Salesforce.
Congratulations to the Cresta team on joining Salesforce.
By
Long before generative AI became a household term, Zayd Enam was chasing a practical question: Could artificial intelligence help people become better at their jobs?
That question became Cresta. Turning it into the company that’s now joining Salesforce took years of getting close to customers, proving the product, and building a team that could deliver.
Cresta began in 2017 out of the Stanford Artificial Intelligence Laboratory. Zayd was the driving founder, bringing together a founding group that included Tim Shi, Sebastian Thrun, and Andre Esteva. Their thesis was simple, but contrarian for its time: build AI around people.
Customer service was a natural place to start and offered an enormous opportunity. Anyone who has struggled through an automated phone tree understands the shortcomings of rigid scripts and chatbots. Cresta studied what the best human agents did well and helped their colleagues do the same. Its technology analyzed customer conversations and delivered guidance while the interaction was still underway. An agent could receive a suggested response, help handling an objection, or a prompt to ask a better question.
Co-founder Tim Shi remembers the print of Diego Rivera’s Detroit Industry that Zayd hung in their first Redwood City office: workers with every tool arranged within reach. They wanted the same thing for knowledge work: the right answer at hand, the moment a rep needed it. Cresta was putting generative AI to work years before the category became mainstream.
Cresta emerged from stealth in 2020, just as businesses were shifting to remote work. Managers could no longer sit beside a new employee or coach across a contact center floor. They needed a way to train and support distributed teams. Cresta’s ability to deliver guidance in the moment became critical. Adoption accelerated. Product-market fit was taking hold.
From proving the product to scaling the company
Zayd had also recruited an exceptional engineering team. Many of those early engineers remain at Cresta today, carrying forward years of technical knowledge and customer understanding. One of his most consequential recruits was Ping Wu, who had co-founded Google’s Contact Center AI effort. Ping joined Cresta in 2021 and became CEO in May 2023, succeeding Zayd. He brought deep technical expertise and the operating experience to scale the business for demanding enterprise customers.
Under Ping, Cresta strengthened its management team, sharpened execution, and expanded the product. Building on its foundation of assisting human agents, the company introduced Cresta AI Agent, capable of handling complete customer interactions. Enterprises could now use Cresta to decide where AI could solve a customer’s problem directly and where a person could deliver the best outcome.
Delivering on that promise took more than better language models. Customers needed products they could rely on, integration with the systems they already used, and people accountable for results. Ping and his colleagues built the organization to deliver all three while continuing to push the technology forward.
Salesforce’s global reach and world-class platform give Cresta the opportunity to bring its technology to many more enterprises. Inside Salesforce, the team will continue transforming customer service.
Cresta’s journey shows what it takes to build a successful AI company. Research creates possibilities. Founders turn those possibilities into products by getting close to customers. Enduring companies then require leaders and teams who can deliver those products reliably, at scale. Zayd supplied the conviction and persistence to get Cresta started. Ping built on that foundation and led its next chapter of growth. Across both chapters, the team kept translating advances in AI into practical value for customers.
Congratulations to Zayd, Ping, the founders, and everyone who helped build Cresta. From a Stanford research project to mission-critical business infrastructure, you showed what it takes to put AI to work.