A Generative Artificial Intelligence Engineer develops, designs, and manages generative models and algorithms that support the generation of new content in the form of images, text, audio, and other multimedia. They utilize GPTs, GANs, VAEs, and other deep learning architectures to craft systems capable of generating data. May work with data scientists, machine learning engineers, and software developers.
Senior Director, Applied AI at Otterbrook Senior Director, Applied AI at Otterbrook in Campbell, California Posted in 5 days ago.
Type:
full-time
Job Description:
Senior Director, Applied AI Location:
New York or San Francisco, hybrid
Openings:
2
Compensation:
$220,000 base plus significant fund equity Search run by: Otterbrook | Confidential, client name shared on a first call The Firm A venture and applied technology firm that builds and backs healthcare companies.
Two strategies:
venture capital, backing founders at the earliest stage, often before the idea is fully formed, and venture buyout, taking control positions in proven companies and rewiring their products with AI. The portfolio spans a wide range of healthcare, from payments to clinical technology. They also hold research partnerships with major health systems that give them access to clinically validated problems. The team A small, senior applied AI practice sitting inside the firm. Former founders, architects, and research scientists. They have been working on language models since before they were large. The practice owns the firm's technical investments, supports founders with architect-level guidance, and builds the agentic operating system the business runs on. The role The firm invests in AI founders and is now putting itself through the same transformation. These two hires build the applications and agentic workflows that power the firm's own decisions.
In practice:
Which markets to expand into, and which thematic areas to invest behind Which concepts to generate and pursue Which founders to back, and how to score and rank them How the firm sources, assesses, and hires talent You embed directly with the investment and talent teams, run discovery on loosely defined problems, and ship working software into their hands. Roughly 99% of the work is internal to the firm. Systems that work well may get deployed to portfolio companies or to outside investors. This is an individual contributor role. Hiring and managing engineers is possible over time, not on day one. You report to the VP of Applied AI, who built the firm's virtual investment committee himself and published on the agentic harness behind it. He is a player-coach with a long track record leading data and AI organizations at major financial institutions, and he still writes code. What they're screening for This is not a title search. Strong candidates have been forward deployed engineers, applied engineers, ML engineers, founding engineers, or technical founders. The shape of the work matters, not the label. You ship applications, not pipelines. Products real people use. If the work has been mostly data engineering, modeling, or infrastructure with no product surface, this is not the fit. You have rebuilt your own workflow around agentic coding. Deep in Claude Code or equivalent, with your own skills, MCP servers, and custom harnesses. This is a hard screen. They have interviewed a lot of people who talk about AI and few who have actually retooled how they work. You have run agentic systems in production. Beyond basic RAG. You have dealt with what breaks when models sit in front of real business users: evals, context engineering, memory, tool design, failure recovery. You are hands-on today. Not someone who used to build. The most common reason candidates get passed on here is being more of a people leader than a builder. You are strong on both sides of the line. You can sit across from an investor or a partner, pull out what they actually need rather than what they asked for, and ship it. Senior nontechnical people trust you because you deliver. You operate in ambiguity. You define the problem as often as you solve it. Not required Healthcare experience. Explicitly not a screen. A specific number of years. Titled senior because they want someone who ramps without hand-holding, but they are opportunistic on level. Depth of what you have shipped counts for more than years on paper. Strong plus You have built decision-support software for investors, bankers, or dealmakers. Venture, private markets, hedge fund, or investment banking context all count Prior founding experience, particularly 0 to 1 A pre-LLM background in ML, knowledge graphs, or semantic search CRM-style intelligence products such as RevOps or recruiting software Compensation $220,000 base, with flex for the right person. The comp philosophy is deliberately modest cash and significant equity. The equity is fund participation rather than standard startup options, on a six year vest with a guaranteed annual vest. The firm walks candidates through the structure directly on a first call, since it is unusual enough that a summary does not do it justice. Full benefits. Hybrid from the New York or San Francisco office. Process Four rounds, three virtual and one onsite. First conversation is with the hiring manager, then internal talent, then leadership. A build exercise sits in the middle of the process. Expect something practical and close to the real work rather than an algorithm test. They are not expecting a finished product, they want to see how you think and what you reach for. They have committed to moving at candidate pace rather than dragging it out. How they work Small, senior, low ego. Rigorous debate is expected and so is shipping. Their internal standard is that AI sets the floor, and taste and judgment own the last mile
