The FDE Manifesto

The most valuable engineer in the room is no longer the one who writes the most code.

It's the one who can walk into an organization they don't work for, understand what is actually broken, and make software land inside it.

We're building the commons for those people.

The problem we're here to fix

The forward deployed engineer is one of the most sought-after roles in software, and nobody agrees on what it is.

Search for it and you'll find a dozen guides published this year by companies selling something adjacent. No two define the role the same way. There's no ladder, no shared vocabulary, no vendor-neutral institution, and nowhere for the people actually doing the work to find each other.

So an engineer with eight years of experience can't tell whether she already qualifies. A VP told to stand up an FDE function invents the job description from scratch. Both are solving the same missing-standard problem alone, in parallel, badly.

That's a coordination failure, and coordination failures are exactly what communities are for.

What we believe

The role is real, and it's not a rebranded consultant. A product engineer builds one capability for many customers. A forward deployed engineer builds many capabilities for one. That inversion changes the skills, the incentives, and the definition of good work.

The job is organizational before it's technical. The first system you have to read is not the codebase. It's who decides, who is quietly opposed, and which documented process nobody follows. Engineers who skip this fail the role no matter how well they write code.

The value is in what comes back. An FDE who only delivers for one customer is an expensive contractor. The model pays for itself when field reality returns as product direction. That loop is a competency, and it belongs in the definition.

This work gets more important, not less. As AI collapses the cost of producing software, the bottleneck moves to deploying it into messy human organizations with legacy data, political stakeholders, and people who will not change their workflow because a model told them to. That difficulty is organizational, not computational, and it's not on the same curve. The forward deployed engineer is the role that absorbs the value as the ones around it compress.

A standard nobody can fork isn't a standard. It's marketing. Ours lives in a public repository under an open licence, with a changelog and an argument history. Disagree with it in a pull request.

What we refuse

We will not sell you a certificate. The framework is free, permanently. The moment a definition is something you have to buy, it stops being a commons and starts being a product with a conflict of interest.

We will not be a vendor's community. Not Palantir's, not any lab's, not Rohith's. No single company owns the definition of a role that many companies hire for.

We will not fake a crowd. No invented member counts, no seeded chapters in cities where nobody has volunteered, no placeholder job listings to make the board look busy. On the day we launch, most of this is empty, and we would rather show you that than stage it.

We will not pretend the guides are worthless. Some of the existing writing on this role is genuinely good. Where it is, we will link to it and say so, including when we didn't write it.

We will not gatekeep the title. The framework is a mirror, not a bouncer. It exists so people can locate themselves and find the next step, not so anyone can be told they don't qualify.

What we're building

A canonical, versioned competency framework. Learning routes in from the jobs people actually hold today. Writing by people doing the work, not people selling to them. Local chapters run by local organizers with a real playbook and real support. Eventually, country events and one global gathering. A job board, once there are people on it worth hiring.

In that order, because that's the order in which each thing becomes honest.

How this goes wrong

The most likely failure is that this becomes another content site: a lot of published words, no practitioners, no argument, no correction. A standard with no dissent in its history is a standard nobody used.

The second most likely failure is that it becomes a job board with a manifesto attached.

We would rather have fifty people who argue with the framework than five thousand who skim it.

The invitation

If you do this work, you already know things that aren't written down anywhere. The framework in this repository is a first draft by one person with eighteen years of context and a strong opinion. It's wrong in places. Finding out where is the entire point.

Read it. Locate yourself on it. Then tell us what it got wrong.