a16z Launches Ops Engineering Fellowship Program
Andreessen Horowitz ecosystem partner David Booth announced the launch of the a16z Ops Engineering Fellowship, gathering about 50 internal deployment engineers and "AI-immersed" operators in San Francisco and New York for six weeks, with group communication included.
Booth defines operations as a field that must be approached with engineering thinking. He sees a new class of builders within companies connecting systems, building agents, and redesigning finance, sales, customer service, and intermediary processes. Box CEO Aaron Levie refers to them as "frontline engineers for internal deployment," while others call them AI operations or agent engineers, with no unified title yet. Booth believes that because the terminology is not settled, it is necessary to bring the leading individuals to the same table.
Announced founding members include: Claire Vo (founder of chatPRD, serving as Founding Fellow), Sebastien Goddijn (formerly of Ramp, working on internal product Glass), Owen Williams from Stripe, Ann Miura-Ko (co-founder of Floodgate), Brandon Gell (COO of Every), TK Kong (involved in launching Ramp Labs), and Matt Portnov (head of internal AI at Decagon). Katie Kirsch's list also includes Ben Tossell. Booth mentioned that several others could not be publicly listed yet.
This is not a16z's first career Fellowship. Earlier this year, the organization held eight-week batches for frontline deployment engineers, design engineers, growth engineers, and talent engineers, with scales typically ranging from 50 to 70 participants selected from thousands of applications, with members from companies like OpenAI, Ramp, Decagon, and Cursor. The operations engineering batch is written as six weeks, expanding the location to New York and shifting the focus from frontline engineers delivering externally to those internally integrating agents into finance and customer service.
Booth is publicly soliciting two lists: individuals truly responsible for "operations engineering" within companies, and industry figures who should be invited to this session or dinner. The application channel and details page will be released simultaneously. The debate over titles is seen as a signal: the roles have emerged, but the HR systems have yet to provide standard job levels.
In market mechanisms, this is a warm-up for talent networks and project flows, not an immediate buy-sell situation. Buyers are companies looking to hire internal agent engineers and those selling enterprise workflow software; sellers are traditional functions that still treat operations as process documentation and outsourced seats. Funding will not flow on the day of the Fellowship, but the list will determine who sees the next batch of internal tool entrepreneurs first and who can be poached by combination companies. Beneficiaries include companies like Ramp, Decagon, and Stripe that already have internal AI teams, while those who only sell generic chatbots and do not modify customer internal systems will be under pressure.
On a supplementary level, Booth had already described this role months ago as a pipeline for engineers to business operations, or frontline engineers internally, also mentioning chief of staff and "IT leaders who think strategically." The frontline deployment model pioneered by Palantir has been turned into billion-dollar external organizations by companies like Amazon Web Services; a16z is now mirroring the same persona within company walls.
Source: Public Information
ABAB AI Insight
a16z is establishing the new profession as a Fellowship to seize the definition during its formation period. David Booth, coming from On Deck and AngelList, excels at gathering scattered individuals into a network that can be tapped by the fund. The frontline deployment engineer batch targets those delivering to clients; the operations engineering batch focuses on those within client companies who modify processes. Both sides stem from a Palantir-style embedding: the only difference is whether the badge hangs on the seller or the buyer.
The capital path is to first gather people, then circle the next generation of internal tool companies. Six weeks of local dinners and group chats will expose practical operations like automated financial reconciliation, customer service agent orchestration, and sales system connections; these pain points are seed projects. The appearance of Ramp Labs, Decagon's internal AI, and Every operations on the same founding list indicates that the fund aims to transform "operational agents" from hidden roles into an investable track. Ann Miura-Ko from Floodgate joining also integrates operator training into the same network.
A similar structure is a16z's replication for growth engineers and talent engineers: each emerging job level first holds a session, and the list becomes a map. The industry is at a window of external FDE supply shortage and internal roles lacking standard titles. Amazon Web Services has already invested billion-dollar-level internal resources for external FDE, indicating that the delivery end has been productized; the internal end is still at the stage of "no name settled," and whoever first organizes a class will gain the recommendation power and hiring standards.
Essentially, this is a reconstruction of the industry chain. The mechanism is: after model capabilities spill over, the bottleneck shifts from algorithms to whether internal company systems can be connected; the individuals connecting them simultaneously grasp the truth of processes and purchasing power. By organizing these individuals into a concurrent network, the fund gains a project radar more accurate than a business card, while startups acquire a batch of buyer engineers who can implement solutions themselves.
ABAB News · Cognitive Laws
When titles are not yet settled, the network is already worth a fund cycle.
Engineers delivering externally sell systems, while engineers connecting internally modify companies.
Whoever first organizes education for new roles will set the pricing for those roles.