UPDATE: OpenAI’s head of forward-deployed engineering, Colin Jarvis, has revealed groundbreaking strategies for transforming AI hype into tangible results for major companies. In a recent episode of the Altimeter Capital podcast, aired on July 13, 2023, Jarvis explained how his small team of 39 engineers is driving significant value in the AI landscape, generating “tens of millions to sometimes the low billions” for clients.
Jarvis emphasized that the team’s approach involves embedding engineers directly with clients, a model gaining traction in the tech industry as organizations struggle to translate AI enthusiasm into effective applications. “People were really excited [when ChatGPT launched], but it was kind of hard to get value from the models,” Jarvis noted. The forward-deployed team aims to bridge this gap by learning workflows and collaborating closely with client staff.
One of their standout projects involved Morgan Stanley, an early adopter of OpenAI’s GPT-4. The technical setup took merely six to eight weeks, but gaining the trust of financial advisors required an additional four months of pilot testing and evaluations. Ultimately, Jarvis reported that about 98% of the advisors integrated the technology into their operations.
The forward-deployed engineering model, originally popularized by Palantir, is now pivotal for OpenAI’s strategy. Jarvis stated that they aim to avoid “services revenue” and focus on creating product playbooks that provide long-term value to customers. OpenAI has ambitious plans to expand the team to 52 engineers by year-end, with 24 job openings available across the US, Europe, and Japan, offering salaries up to $345,000 plus equity.
In addition to financial services, the team has collaborated with a European semiconductor company to develop a “debug investigation and triage agent.” This solution addresses a significant pain point, as engineers reportedly spent 70% to 80% of their time debugging chips.
Jarvis’s forward-deployed engineering function is seen as critical to scaling AI technology. Earlier this year, he expressed in a LinkedIn post that the goal is to accelerate customer production, whether through new applications or scaling existing cases. This commitment to innovation was echoed by OpenAI’s international managing director, Oliver Jay, at the Fortune Brainstorm AI 2025 conference, where he highlighted the model’s specificity in advancing AI deployment.
The urgency surrounding AI adoption is palpable. Venture investors are taking note; Diana Hu, a partner at Y Combinator, remarked in a June podcast that founders leveraging forward-deployed engineers are closing “six, seven-figure deals” with major firms, indicating a competitive edge over established giants like Salesforce and Oracle.
As OpenAI continues to recruit and expand its forward-deployed engineering team, the tech community watches closely for how these developments will reshape the landscape of AI integration in the corporate world. Stay tuned for more updates as OpenAI pushes the boundaries of what AI can achieve in real-world applications.
