The Expanding Scope of Forward Deployed Engineering and Its Ties to Outcome-Based Pricing
Natalie Mier, head of agent engineering at Sierra, argued in a recent talk that forward deployed engineering has accumulated responsibilities across four phases—DevOps, data integration, custom solution building, and enablement—so that it now means something different at nearly every company. She contends that with AI agents lowering the cost of code, the role's core accountability to customers is converging with product engineering, and that outcome-based pricing models make this generalist role more central. Mier also noted that the term may lack a coherent definition, but someone still must be accountable for whether software works for the customer.
Listen to this dispatch
Narrated by an AI-generated voice.
At a recent conference talk, Natalie Mier, head of agent engineering at Sierra, began by describing the "dirty secret" of forward deployed engineering: the discipline "doesn't exist." The term covers so many responsibilities that it means something different at nearly every company, yet it has become, in her telling, one of the most sought-after roles in AI. Her argument, traced through her time at Palantir from 2016 to 2021, was less that the job is a myth and more that it has accumulated nearly every skill a software company can ask for.
Mier traced the role's evolution at Palantir. In 2008, forward deployed engineering was effectively DevOps: platform stability for on-premises deployments, with engineers physically sitting at customer sites. By 2012, the job added data integration, as engineers built ontologies that modeled customer data. In 2016, it gained custom solution building, often through Palantir's Slate dashboard tool. After 2020, it included enablement—teaching customers to use the platform themselves, as with Palantir's Airbus work. These weren't sequential replacements, Mier noted; responsibilities stacked. By the time she left, the role had grown to encompass all four phases simultaneously.
That history underpins her broader point: the single continuity across every vintage of FDE is accountability to a customer, whether for uptime, data modeling, or teaching. The rest varies. Today, she argues, that accountability is converging with product engineering. When code becomes cheap to produce with AI agents, the distinguishing work is translating customer outcomes into products—and the lines between engineering disciplines blur.
Mier connected this to a commercial shift she says Sierra has bet on: outcome-based pricing. Agents, she argued, make it possible to price software by the outcome it delivers rather than by seats or usage, because the agent's effect on a customer interaction is relatively attributable. That pricing model, she said, makes the customer-accountable generalist role more central, not less.
The talk was, by her own framing, a set of arguments rather than settled fact. She acknowledged writing a 2024 article proposing "agent engineering" as a distinct discipline with the customer accountability of FDE but its own scope, and said she now sees it as a flavor of FDE rather than a separate field. She cited Google Cloud's recent customer engineering hiring effort and OpenAI's new unit as evidence the role is spreading, and noted a blog post from Sierra's head of go-to-market ops, Elliot Greenwald, on changing software sales models.
Her closing line—"Forward deployed engineering is dead. And long live forward deployed engineering"—captures the ambiguity she sketched: the term may lack a coherent definition, but someone still has to be accountable for whether the software actually works for the customer who bought it. She ended by noting Sierra is hiring for FDE roles.