An OpenAI exec explains how his growing team helps companies move from AI hype to adoption

At OpenAI, a specialized team is making strides in the integration of artificial intelligence within major corporations. Led by Colin Jarvis, the forward-deployed engineering team works directly with clients to transform AI models from abstract concepts into practical applications that generate significant financial returns. Currently comprised of 39 engineers, the team is set to expand to 52 by the end of the year. OpenAI is actively recruiting for this team with 24 positions available across the US, Europe, and Japan, offering competitive salaries that can reach $345,000, along with equity options. The term 'forward-deployed engineer' has gained recognition in the industry, particularly from companies like Palantir, and refers to engineers who collaborate closely with clients to tailor products to their specific needs. Jarvis notes that the initial excitement surrounding AI models, particularly after the launch of ChatGPT in 2022, often did not translate into tangible value for early enterprise users. Many companies found it challenging to leverage these technologies effectively. To bridge this gap, OpenAI adopted a hands-on approach, embedding engineers within client teams to understand their workflows and develop solutions collaboratively. One notable collaboration was with Morgan Stanley, where OpenAI's team facilitated the deployment of GPT-4. While setting up the technical infrastructure took only six to eight weeks, gaining the trust of financial advisors proved to be a more prolonged process. After several months of pilot programs and feedback sessions, Jarvis reported that nearly all advisors—about 98%—ended up embracing the technology. Additionally, the team partnered with a semiconductor firm in Europe to design a 'debug investigation and triage agent', aimed at addressing the extensive debugging time engineers were dedicating—often up to 80% of their work hours. Jarvis emphasizes that the team’s mission is to focus on creating product playbooks rather than generating service revenue, ensuring they stay centered on their goals. Earlier this year, Jarvis shared his vision for the forward-deployed engineering role on LinkedIn, stating that the objective is to expedite clients’ journey to production, whether by pioneering new applications of technology or scaling existing successes. With ongoing recruitment efforts in cities such as San Francisco, New York, and London, OpenAI is committed to reinforcing this innovative approach. The forward-deployed engineering model is being recognized by venture capitalists as well. Diana Hu, a partner at Y Combinator, highlighted on a recent podcast that founders have successfully secured substantial contracts with major corporations through this model. Y Combinator CEO Garry Tan also acknowledged that this strategy offers AI startups a significant advantage against established players like Salesforce and Oracle.