On-Site vs Remote FDE Work: Travel Realities, Per Diems, and Client Expectations

The Forward Deployed Engineer (FDE) role sits at the sharp end of the spear. You aren't just writing code in a vacuum; you are embedding within a customer's technical environment to unblock million-dollar contracts. The "where" of this work isn't a lifestyle perk—it's a fundamental architectural decision about how trust, context, and velocity are managed.

We'll cut through the generic "WFH vs RTO" noise to look at the specific trade-offs of on-site vs remote FDE work, covering travel frequency, per-diem economics, air-gapped realities, and why your physical location directly impacts your career leverage.

The Core Split: Builder vs. Deployer

Unlike standard software engineering where remote work often means isolated ticket-pulling, FDE work has a dual nature. There is a "Builder" phase and a "Deployer" phase.

• Builder Phase (Remote-First): You are writing integration code, building custom data pipelines, or forking a React dashboard. This looks like standard remote dev work. You might be building a multi-agent research assistant locally to prototype a solution for a client.
• Deployer Phase (On-Site Mandatory): You are pushing that code to a Kubernetes cluster that sits behind a client's firewall in a SCIF. You cannot do this from a WeWork. You must physically travel.

Most FDE roles split roughly 60/40 or 70/30 between remote building and on-site deploying, but this varies wildly by sector.

The On-Site Reality: 4 AM Flights and Sensitive Air-gapped Networks

When a client pays a seven-figure annual contract, they don't want a Zoom call. They want you in the room. On-site FDE work isn't just about shaking hands; it's about physics.

The Air-Gap Scenario

Many FDEs work with defense, finance, or critical infrastructure clients. These environments are "air-gapped"—no internet connection. You cannot SSH in. You must carry a "sneakernet" device (a ruggedized laptop with hardware encryption) through physical security checkpoints.

An on-site week looks like this:

• Monday 4 AM: Flight to the client site.
• Monday 10 AM: Clear security, set up a temporary dev environment on a disconnected server.
• Tuesday: Debug a kernel panic caused by a proprietary OS patch that only exists in this building. You cannot Google the error code. You must rely entirely on local man pages and your own brain.
• Wednesday: Whiteboard session with the client's chief architect who refuses to use digital whiteboards due to security policy.
• Thursday: Live fire demo.
• Friday: Red-eye flight home.

Travel Cadence

In a high-touch enterprise role, expect 50-75% travel. This doesn't mean 2 weeks a month; it often means a brutal cadence of 3 days on-site every single week. You live in airline lounges. You optimize your life around TSA PreCheck and packing cubes.

The Remote Reality: High-Trust, High-Autonomy Async Execution

Remote FDE work is not a junior role. Because you can't tap a client on the shoulder, you must operate with extreme written clarity. The remote FDE's primary tool isn't VSCode; it's the Technical Design Document.

The Async Handshake

When remote, you cannot simply "read the room" to see if a client is confused. You must anticipate failure modes. A remote FDE deploys a solution by shipping a detailed runbook alongside the code.

A remote deployment flow often looks like this:

Notice the lack of "FDE flies to site" in that flow. The remote FDE replaces physical presence with high-fidelity async artifacts. You must be exceptional at writing customer-facing technical docs because a typo in a command block can cause a Sev-1 outage that you can't physically see.

Trust and "The Suitcase Factor"

There is an ugly truth: some enterprise clients believe remote workers aren't "really working." In an on-site engagement, the client sees you suffering in their beige conference room and feels they are getting their money's worth. Remote FDEs must manufacture this trust through over-communication. If you are remote, you need to show velocity: daily summary PRs, screenshots of dashboards, and proactive status updates.

Compensation and Per Diem Economics

The comp structure differs significantly between on-site and remote-heavy roles.

The Per Diem Trap

On-site FDEs often fall into a lifestyle trap. They bank large per-diem checks (often paid out regardless of spending) but blow it on airport beers and expensive dinners to cope with the loneliness of the road. Smart FDEs treat per diems as tax-free bonus income. They meal-prep in hotel rooms with Instant Pots (seriously) and pocket the difference.

The Hybrid Sweet Spot and Career Trajectory

Pure remote FDE work is rare at the entry level. You usually must "earn the remote." The typical career arc:

1. Junior FDE (0-2 years): 80% on-site. You are learning the customer's pain points by sitting next to them. You are a sponge.
2. Mid-Level FDE (2-5 years): 40% on-site. You fly in for kickoffs, critical debug sessions, and QBRs. You build the rest remotely. You are leveraging tools like vLLM for high-throughput inference from your home office to prototype faster than the client expects.
3. Senior/Principal FDE (5+ years): 10% on-site. You parachute in only when the deal is at risk or the architecture is failing. You spend most of your time designing solutions and reviewing code.

The hybrid model is the true sweet spot. It prevents burnout while maintaining the "white glove" service that justifies the high contract value.

The RTO Threat in 2026

Is remote FDE work dying? Not for top performers. However, the market is correcting. In 2021, anyone could demand remote work. In 2026, remote work is a privilege earned by demonstrating you can close complex technical deals without a manager watching you. If you want to stay remote, you must become a "player-coach" who can influence the client's technical direction purely through the weight of your written proposals.

FAQ: On-Site vs Remote FDE Work

Is it better to work remotely or onsite?

It depends on your career stage. Early career: onsite is better for learning political navigation and reading body language. Late career: remote is better for deep work and avoiding burnout. For the FDE specifically, on-site builds trust faster, but remote scales your impact across more accounts.

Is remote work dying in 2026?

Low-agency remote work is dying. Generic ticket-takers are being forced back to the office. High-agency remote work (the FDE model) is thriving because it's directly tied to revenue. If you are generating $2M+ in expansion revenue, nobody cares where your desk is.

What is the difference between on-site and remote jobs?

On-site jobs exchange physical presence for context. Remote jobs exchange written articulation for context. On-site is synchronous and high-bandwidth (body language, whiteboarding). Remote is asynchronous and relies on high-quality documentation and demos.

Is it remote or on site?

The terminology matters for LinkedIn filters. "On-site" means 5 days a week in a specific office. "Hybrid" means a mix (e.g., 3 days in, 2 days out). "Remote" means location-agnostic, though often with time-zone restrictions. FDE roles usually fall under "Hybrid" or "Remote" with a "Travel Required" tag (often 30-50%).

How do I transition from a pure on-site FDE role to a remote/hybrid one?

Build a reputation for "autonomous execution." The best way to negotiate remote work is to show your manager a client email praising a deployment you did entirely async. If you can prove you can handle a Sev-1 incident without waking up the engineering director, you earn the right to work from wherever you want. For building those technical artifacts, consider automating your workflow with tools like a personal finance categorizer to get your admin life under control so you can focus purely on client delivery.