On-Site vs Remote FDE Work: Travel Realities, Cadences, and Where the Role Is Heading

The FDE Travel Paradox: High-Touch vs High-Flex

The Forward Deployed Engineer role exists because software fails at the boundary layer. APIs don't integrate themselves. Legacy mainframes don't magically emit clean JSON. C-suite stakeholders don't trust a dashboard they haven't seen you debug live.

This creates a fundamental tension in the role: you need deep, uninterrupted focus time to build, but you also need physical proximity to the customer's dirty data and political landscape.

The search results for "on-site vs remote FDE work" are flooded with generic software engineering takes. They miss the point. An FDE is not a standard remote developer. You are a surgical instrument deployed into a specific operational theater. The question isn't "is remote better?" It's "what is the minimum viable presence required to prevent the deal from melting down?"

The Two Modes of FDE Work

To understand the travel reality, you must separate the role into two distinct modes:

The best FDEs aggressively protect Build Mode. The worst FDEs let Deployment Mode bleed into every waking hour and burn out in 18 months.

The Physical Cadence: 3 Days On-Site, 2 Days Remote

The dominant pattern for mid-to-large enterprise FDEs (Palantir, Scale AI, Applied Intuition) is a weekly commute, not a relocation.

The standard rhythm looks like this:

• Monday: Fly out (5am alarm, airport lounge, arrive by noon). Afternoon: kickoff meeting, walk the floor, plug into the customer network.
• Tuesday–Wednesday: Deep on-site work. Whiteboarding sessions, pairing with customer engineers, grabbing dinner with the client's technical lead to understand the real problem they didn't put in the SOW.
• Thursday: Handoff and demo. Show progress, gather feedback, set expectations for remote work. Fly home in the evening.
• Friday: Remote Build Mode. No internal meetings. Process the week's chaos into clean code and architecture decisions.

This is not a "remote" job. It is a high-travel hybrid job. You will spend 50–75% of your working days physically with the customer during active engagements. When you are between deployments or in a maintenance phase, that flips to 90% remote.

The Financial Mechanics of FDE Travel

FDE compensation packages are structured differently from standard remote engineering roles precisely because of this travel burden. If you're evaluating an offer, you need to model the real comp, not just the base salary.

For a deep dive on the numbers, see our breakdown of Forward Deployed Software Engineer OpenAI Salary & Compensation. Here, we focus on the travel-specific mechanics.

Per Diem Arbitrage Most firms don't require receipts for meals during travel. They pay a flat per diem (often $60–$100/day in the US). Smart FDEs pack breakfast from the hotel lounge, eat a modest lunch, and pocket the difference. Over a 40-week travel year, this can add $8,000–$15,000 in tax-free income.

Points and Status You will fly. A lot. A mid-career FDE on a single long-term deployment can easily hit airline Platinum status and hotel Diamond status within 6 months. The personal value of upgrades, lounge access, and free vacations funded by points is a material part of total compensation. Treat your loyalty programs like a second job. Never split airlines unless forced.

The Hidden Tax: Relationship Decay The financial upside has a cost. Missing birthdays, anniversaries, and the slow erosion of your local friend group is the real price. The FDEs who last more than 3 years are either single with a high tolerance for solitude, or have partners who explicitly bought into the lifestyle.

Remote-First FDEs: The Exception That Proves the Rule

There is a small but growing cohort of truly remote FDEs. They work for companies selling to digital-native startups or mid-market tech companies where the customer's CTO is comfortable with a Zoom-first relationship.

These roles look more like a Solutions Architect with a heavier coding load. The travel drops to 10–20%, usually just for quarterly business reviews and the initial kickoff.

The Tradeoff: Remote-first FDE roles typically pay 15–25% less in base salary than high-travel roles at companies like Palantir or Shield AI. The equity is often less generous. You are trading cash and career velocity for geographic stability.

If you're hunting for these roles, our guide on AI Engineer Remote Jobs Worldwide: How to Find & Land Global Roles covers the search strategy, though you'll need to filter specifically for the "deployed" component.

The Tax Trap: Why "Working from Anywhere" Is a Lie

One of the "People Also Ask" questions hits the nail on the head: Where do I pay taxes if I work remotely and travel?

FDEs are uniquely exposed to tax complexity. You are not a digital nomad working from a beach in Bali for a company that doesn't care. You are entering customer facilities, often in different states or countries, and performing work that creates a physical nexus.

The Rules You Will Break If You're Not Careful:

1. State Income Tax: If you work from a customer site in New York for 3 days a week, New York wants its cut. Your home state of Texas (no income tax) won't protect you. You may need to file in multiple states.
2. Permanent Establishment Risk: If you are the sole on-the-ground presence for your company in a foreign country for 6+ months, you might accidentally create a "permanent establishment" for your employer, triggering corporate tax obligations. This is a fast way to get a panicked call from your CFO.
3. The "Hotel as Home" Myth: You cannot simply claim your parents' address in a low-tax state and call it a day. Tax authorities look at where you physically perform the work, not where you sleep.

Practical Defense:

• Use a tax preparer who specializes in multi-state filings. It will cost you $2,000–$4,000/year. Pay it.
• Track your physical location daily. A simple Google Sheet with dates, locations, and work purpose is sufficient evidence in an audit.
• If international travel exceeds 90 consecutive days, proactively email your HR and legal team to flag it.

Tooling for the Transient FDE

Your environment is fragmented. Monday you're in a SCIF with no internet. Tuesday you're on a customer's VDI. Friday you're on your home rig. Your tooling must be stateless and portable.

The Portable Dev Stack:

• Editor: VS Code with Settings Sync enabled. Your extensions, keybindings, and snippets follow you everywhere.
• Terminal: A well-maintained .dotfiles repo that you can clone in 30 seconds onto any fresh machine. Everything lives in tmux sessions.
• Notes: Obsidian with a vault synced via a private GitHub repo. Customer context decays rapidly; write everything down.
• AI Copilot: An LLM that understands your codebase. For complex technical deep-dives while on the road, the pattern in A Working Engineer's Pattern for Using LLMs to Learn Complex Technical Topics is invaluable for ramping on unfamiliar customer tech stacks quickly.
• Communication: A single, high-quality noise-canceling headset (Bose QC or Sony XM series). Never rely on the customer's conference room mic.

The Physical Kit:

• A portable second monitor (15.6" USB-C). The productivity delta between a single laptop screen and dual screens is the difference between hitting a Friday deadline and working Saturday.
• A travel router (GL.iNet). You will encounter hotel Wi-Fi that blocks SSH. Tunnel everything through your own encrypted connection.

Where the Role Is Heading: The Hybrid Default

The pandemic-era fantasy of fully remote FDEs is dying. Customers are back in their offices, and they expect you there too. The question "Is remote work going away in 2026?" has a specific answer for FDEs: pure remote FDE roles are contracting, but the high-travel hybrid model is expanding.

The reason is economic. An FDE who ships a working prototype in 7 days—as described in How FDEs Turn a Messy Customer Problem into a Shipped Prototype in 7 Days—is worth $500K+ in contract value. The $1,500 weekly travel cost is noise. The real risk is not deploying the engineer and losing the deal to a competitor who does.

The emerging model:

• Weeks 1–2: On-site, high-intensity. Live in the customer's problem.
• Weeks 3–8: Remote Build Mode, with 2 days on-site every other week for check-ins.
• Week 9+: Transition to internal team, FDE moves to next deployment.

This cadence maximizes the FDE's impact while giving them enough remote time to actually build. The companies that force 100% on-site for 6+ months are burning through talent and paying for it in attrition.

FAQ

Is it better to work remotely or onsite as an FDE? It's not a binary choice. The role is inherently hybrid. Full remote means you're not doing the "deployed" part of the job. Full on-site leads to burnout and sloppy code. The sweet spot is 50–70% on-site during active engagements, with protected remote Build Mode days.

Is remote work going away in 2026 for FDEs? Pure remote FDE roles are becoming rarer. Companies have realized that the first two weeks of a deployment require physical presence to build trust and discover the real requirements. However, the 5-days-a-week-in-a-customer-office model is also dying. The winning model is high-travel hybrid.

How to make $1000 a week remotely as an FDE? If you're a full-time FDE, $1,000/week is far below market rate. Entry-level FDE roles start at $120K+ base. If you're asking about side income, the most natural path is converting your deployment scripts and internal tools into a productized consulting offering for smaller firms in the same vertical. But beware of non-compete clauses.

Where do I pay taxes if I work remotely and travel as an FDE? You pay taxes where you physically perform the work. If you spend 3 days/week in California at a customer site, you owe California income tax for those days. If your employer is in a different state, you likely need to file in both. International deployments add another layer. Hire a specialized multi-state tax accountant before you start, not after the audit letter arrives.