San Francisco: A decision to put college on hold and join OpenAI changed the trajectory of Yash Patil’s career.
A few years later, the Stanford student-turned-entrepreneur is preparing to take on a senior AI leadership role at Palo Alto Networks after the cybersecurity company agreed to acquire his startup, Applied Compute, for $4.5 billion.
Yash Patil, who co-founded Applied Compute in May 2025 with Rhythm Garg and Linden Li, will become Co-Chief AI Officer at Palo Alto Networks once the acquisition is completed.
The all-stock transaction, announced on November 18, is expected to close in the company’s fiscal third quarter of 2027, subject to customary closing conditions and regulatory approvals.
The deal represents a remarkable transition for Yash Patil, whose journey took him from university life to one of the world’s most prominent AI companies, and eventually to building a startup that reached a valuation of $3.25 billion before agreeing to be acquired.
It also highlights how an early career decision, followed by a willingness to build a company around an emerging technology opportunity, can lead to an unexpected entrepreneurial outcome.
From Stanford University to OpenAI: The Career Decision That Changed Yash Patil’s Life
According to his personal website, Yash Patil took leave from Stanford University to join OpenAI as a Member of Technical Staff. His LinkedIn profile records his tenure at the AI company from October 2023 to May 2025.
Rather than following the conventional path of completing university before entering the technology industry, Yash Patil chose to work directly with one of the companies at the forefront of artificial intelligence development.
At OpenAI, he became a key contributor to Codex, the company’s agentic software engineering initiative, according to Applied Compute’s launch announcement.
That experience brought him into an environment where AI models were evolving beyond conversational tools towards systems capable of performing complex tasks.
His eventual co-founders also came from OpenAI. Rhythm Garg was a core contributor to o1, the company’s first RL-trained reasoning model, while Linden Li worked on machine learning systems and infrastructure for reinforcement learning training.
The three eventually decided to build a company of their own.
The Leap From Working at OpenAI to Building an AI Startup
In May 2025, Yash Patil, Garg and Li founded Applied Compute.
The startup was built around a belief that companies would increasingly need to own and improve their own AI capabilities instead of depending entirely on externally developed models.
When Applied Compute publicly launched on October 29, 2025, its founders outlined their argument in a note titled “It’s time to get specific”.
They believed competitive advantage would come from AI agents designed for the requirements of individual businesses, using models tailored to their operations.
The company entered the market with $80 million in backing from investors including Benchmark, Sequoia, Lux, Hanabi, Neo, Definition and Elad Gil.
Its early customers included Cognition, DoorDash and Mercor.
The decision to build an enterprise AI infrastructure company quickly attracted investor interest. In April 2026, Applied Compute secured another $80 million at a post-money valuation of $1.3 billion, taking its total funding to $160 million.
In August, it introduced AC2, the Applied Compute Agent Cloud, in private beta. The platform was designed to bring together AI model training, inference and continuous improvement, allowing businesses to operate models they own.
From Zero Revenue to $100 Million in 18 Months
The scale of Applied Compute’s growth became clearer in November 2026.
On November 2, Patil and his co-founders announced that the company had crossed an annualised revenue run-rate of $100 million, having started from zero revenue approximately 18 months earlier.
The company also raised $350 million at a $3.25 billion valuation in a round led by Elad Gil. The financing included participation from Kleiner Perkins, Lux Capital, Hanabi Capital, NVIDIA, Blackstone, Shopify Ventures and HubSpot Ventures, among others.
Applied Compute reported that its inference infrastructure was processing more than 1 trillion tokens daily. Customers had conducted over 1,700 training runs on AC2 since July, while 40% of its customers used the platform for both training and inference.
Its customer base had expanded to include Microsoft, Shopify, Mercor, Harvey, DoorDash, Cognition, Clay and Decagon.
The company also said its workforce had tripled since January, with researchers and engineers accounting for 85% of its employees.
On November 12, Applied Compute acquired Delegate, a startup founded by former Scale AI employees Felix Su and Jason Yang.
Six days later, the company announced the agreement that would take Patil’s entrepreneurial journey into a new phase.
A $4.5 Billion Deal Changes the Next Chapter
On November 18, Palo Alto Networks announced its agreement to acquire Applied Compute for $4.5 billion in an all-stock transaction.
Following the acquisition, Patil will become Co-Chief AI Officer, while the Applied Compute team will join the cybersecurity company.
Announcing the deal on X, Yash Patil reflected on the original idea behind Applied Compute.
“Rhythm, Linden, and I founded Applied Compute with a belief that was unusual at the time: that every company should own its intelligence. We think training, serving, and improving proprietary models will become a normal part of running a company.”
He also explained why security became central to the company’s approach.
AI models deployed within enterprises need access to sensitive information and internal systems. They must also comply with organisational policies and remain explainable, verifiable and auditable.
“This is the reason we’ve always prioritized security, and why joining Palo Alto Networks will help us go further faster,” he said.
Palo Alto Networks Chairman and CEO Nikesh Arora said: “Every enterprise is going to run a mix of open and closed models, and every one of those models needs to be secured. Applied Compute has built the best post-training and inference platform we’ve seen, and its research team is exceptional.”
Why Palo Alto Networks Wants Applied Compute
The acquisition is part of Palo Alto Networks’ strategy to strengthen its position in enterprise AI security.
The company expects businesses to increasingly use a combination of open and closed AI models, customised to their data, cost and operational requirements.
Applied Compute’s technology will allow Palo Alto Networks to help customers train, adapt and deploy open-weight AI models with security integrated into the process. The platform is designed to support deployments in public cloud infrastructure as well as sovereign environments.
Applied Compute will continue serving AI-native companies, digital-native businesses and large enterprises following the transaction.







