TL;DR
Across 160 forward deployed engineer postings from 160 companies (MORT Role Insights, July 2026), employers want a hands-on builder who ships production code in the customer's stack, Python first (71%). Most postings (67%) are about building LLM solutions, and daily use of agentic coding tools like Claude Code is a named requirement. Post-sales delivery (84%), communication (76%), and ownership (69%) beat sales polish. Where experience is stated, the median minimum is 5 years, and 72% of postings mention no degree.
Forward deployed engineer is one of the fastest-spreading job titles in tech, and one of the fuzziest. Almost every company means something slightly different by it. So instead of guessing, we read 160 live forward deployed engineer postings from 160 different companies in July 2026 and structured each one individually. Here is what employers actually ask for, and what a serious candidate should do about it.
What a forward deployed engineer actually does: build, not advise
The single clearest finding is that this is a building job. Across the 160 postings, the work tilts overwhelmingly toward building: 53% are coded mostly-building, 31% balanced, and only 1% mostly-advising. It is not consulting with a laptop. Post-sales delivery appears in 84% of postings versus only 16% for pre-sales.Forward deployed engineers show up after the deal closes to build, not before it to help sell.
They also own the scoping work: requirements discovery appears in 70% of postings and solution scoping in 63%, but the endpoint is shipped code, not a scope-of-work document. As SentinelOne puts it, "Unlike a platform or product engineer working behind a backlog, you will embed with the stakeholders who own a problem, scope it end to end." Twilio is blunter about the deliverable: "Write and ship production code (Python, TypeScript, Node) directly in customer codebases." And Smartsheet frames the seniority of it as "building a function, not filling a role."
"Forward deployed" is more literal than the mythology suggests. The work-model split is close (onsite 37%, hybrid 33%, remote 31%), but onsite embedding at customers appears in 55% of postings, more than the onsite count, meaning many nominally remote roles still require periodic embedding trips. Istari captures the median case: engineers who are "embedded directly inside our customer organizations, operating as part of their team while remaining tightly connected to [our] product."
What the postings ask for, in numbers
Here is the share of the 160 July 2026 postings that name each signal. Every figure is a share of the full sample of postings.
| Signal | Share of the 160 postings |
|---|---|
| Post-sales delivery | 84% |
| No education requirement mentioned | 72% |
| Python | 71% |
| Requirements discovery | 70% |
| Builds LLM solutions for customers | 67% |
| Solution scoping | 63% |
| Mentions travel | 60% |
| Senior or staff-plus seniority | 56% |
| Onsite embedding at customers | 55% |
| Junior | 7.5% |
The technical bar: Python, cloud breadth, and LLM apps
Python is the price of entry, appearing in 71% of postings (114 of 160), well ahead of TypeScript (29%), JavaScript (20%), Java (14%), SQL (13%), and Go (9%). Where Python is not primary, the role usually sits on a specialized stack instead. On infrastructure the pattern is breadth over depth: AWS (51 postings), Google Cloud (42), and Azure (38) are usually listed together, with Kubernetes (33) and Docker (27) forming the next tier and Terraform (15) reading as baseline DevOps literacy.
LLM work is the job, not a bonus. 86% of the postings touch AI in some form, and 67% (107 of 160) are specifically about building LLM solutions for customers. The working vocabulary is concrete: agents (59 mentions), RAG (35), prompt engineering (about 28), and MCP (9), with LangChain (12), LangGraph (7), and CrewAI (6) as the named frameworks. Okta's staff posting spells out the breadth it wants: "build production integrations with Claude, ChatGPT, Microsoft Copilot, Agentforce, Bedrock, LangChain, CrewAI, the OpenAI Agents SDK, or MCP servers."
Across all of this, postings draw a hard line at production code, not prototypes. ShipBob wants "Not demos, not proofs of concept, but things that run." GitLab holds the bar even at staff level, requiring forward deployed engineers to "contribute production-quality code" rather than sliding into advisory work as they get more senior.
The sharpest new signal: agentic coding tools as their own requirement
A distinct 2026 pattern in the data is fluency with AI coding tools listed as its own line-item requirement, separate from building AI products. Sixfold asks for candidates where "Agentic coding is your default workflow using Claude Code or Codex." Tread wants "Daily, opinionated use of AI coding tools." This shows up even on roles where AI is not the product.
What this means for candidates
If you are targeting these roles, using an agentic coding tool like Claude Code, Cursor, or Codex should be a daily habit you can talk about concretely, not a demo you tried once. Be ready to describe production software you actually shipped with one.
Experience, seniority, and education
Where a minimum is stated (in 106 of the 160 postings), the median is 5 years, and the distribution skews senior: 56% of postings are senior or staff-plus versus only 7.5% junior. Past staff level, people and delivery leadership layers on top of technical depth rather than replacing it.
Credentials matter less than results. 72% of postings mention no education requirement at all. Zipline states the philosophy directly: "We do not need specific degrees; we need results." Degree requirements cluster at the two extremes, junior pipelines and deep-specialization principal roles, but for most postings production experience substitutes for a degree.
The human side: delivery and ownership beat sales polish
The customer-facing ranking is unambiguous: post-sales delivery (84%), requirements discovery (70%), and solution scoping (63%) dominate, while pre-sales (16%), training or enablement (10%), and executive communication (6%) trail far behind. The data does not support the forward-deployed-engineer-as-executive-whisperer image. Companies want someone who ships in the client's stack, often with no support layer to lean on. Automat spells out the reality: "You own your accounts end to end, with no dedicated success manager, account lead, or hand-off layer standing between you and the customer." GitLab adds the internal version, coordinating "across Product, Engineering, Support, SRE teams without formal authority."
On soft skills, communication (76%) and ownership (69%) dwarf everything else, with problem-solving variants (about 46%) and ambiguity tolerance (about 28%) next, and leadership and curiosity minority signals at about 10% each. Venncity captures the theme: "You've scoped and built products from 0 to 1 in ambiguous environments."
As for who gets hired, the dominant feeder background is software engineer (about 55%), then solutions engineer or architect (about 40%) and consultant (29%), with data scientist, data engineer, and ML engineer forming a secondary tier. If you are coming from the solutions side, it is worth seeing what that market looks like on its own terms in our companion analysis of what employers look for in solutions engineers.
Travel, embedding, and clearance
Travel and embedding are real but uneven. 60% of postings mention travel and 55% involve onsite embedding, though 40% do not mention travel at all (absence of a signal in a thin posting, not confirmation of zero). The range is wide: about 25 to 30% travel at the commuter end (Sigma, DigitalOcean), 50 to 75% of extended periods at customer sites (Solve.ai), 80 to 100% embedding in classified environments (Yurtsai), and outright relocation at the extreme (Sandtech asks for 12 to 24 months inside Tanzania's Ministry of Health). A handful add unusual physical demands, such as Anduril, whose posting involves climbing towers and lifting 50 to 70 lbs.
Security clearance is a loud minority. 87.5% of postings have no clearance requirement; the 20 that do cluster tightly in defense (18) and government (5) roles, with real citizenship gates. Defense and government forward deployed engineering is a recognizable, well-compensated niche, about one in nine postings, not the typical job. Healthcare (13) and financial services (about 15) are the next-largest verticals.
What forward deployed engineers get paid
Salary comes from the postings that advertise a range, so read it as a partial picture rather than the whole market.
| Statistic | Value |
|---|---|
| Median advertised band | $152,000 to $225,000 |
| Median band midpoint | $190,000 |
| Quartile midpoints | $165k / $190k / $221k |
| Top of market | xAI $180,000 to $440,000, Handshake $300,000 to $375,000, Airtable $248,000 to $351,000, Didi (Staff/Principal) $255,000 to $351,000 |
From the 59 postings that state a yearly USD range. A junior anchor sits well below the band (C3 AI, $120,000 to $158,000 for a 2-plus-year role in NYC), and UK and EU postings run materially lower (V7 Labs, £70,000 to £90,000). Only about 37% of postings advertise salary at all, and the sample skews US and larger companies, so treat these as US-market bands.
The candidate playbook
If you want to clear the bar at most of these companies, the evidence points to a specific short list. Every item below is drawn from the same July 2026 postings.
Ship production code, not prototypes
Mostly-building outnumbers mostly-advising 85 to 2. Have an end-to-end production artifact you wrote and deployed, not a proof of concept. "I designed the architecture" is weaker than "I wrote and deployed the code that is running."
Make an agentic coding tool your daily workflow
Claude Code, Cursor, and Codex are named as line-item requirements by employers from startups (Zenity, Rely Health, Zip) to consultancies (Deloitte, Provectus). Be ready to describe production work you shipped with one.
Go deep on Python and LLM application skills
Python appears in 71% of postings and 67% are about building LLM solutions. Learn RAG and agent orchestration concretely: agents (37% of postings), RAG (22%), and the LangChain, LangGraph, and CrewAI family are the default vocabulary.
Get breadth across the big-three clouds
AWS (51 postings), Google Cloud (42), and Azure (38) are listed interchangeably, with Kubernetes, Docker, and Terraform in the next tier. Cloud-agnostic deployment fluency beats single-cloud certification.
Lead with delivery and ownership, not sales polish
Post-sales delivery (84%) and requirements discovery (70%) dominate, while executive communication is only 6%. Communication (76%) and ownership (69%) are the two most-cited soft skills by a wide margin, so practice narrating stories where you owned an account end to end with no handoff layer.
Pick a vertical and go deep if you can
55% of postings name a vertical, and there domain fluency is a second technical bar, not a soft preference (project44 wants freight procurement knowledge, FloQast prefers ACA or ACCA accounting credentials, Cognite wants Oil and Gas or Pharma experience).
Do not let a missing degree stop you
72% of postings never mention education, and production experience substitutes directly. Zipline puts it plainly: "We do not need specific degrees; we need results."
Be honest about travel tolerance before you apply
60% of postings mention travel, and several involve genuinely unusual demands (Normal Computing asks for 12 months onsite in Korea; PhysicsX lists 3 to 4 weeks per quarter globally under ITAR export controls).
Aim at the seniority reality
With only 7.5% junior postings, the realistic entry paths are junior pipelines at big platforms (C3 AI, Cresta, Palantir internships), or lateral entry from solutions engineering, consulting, or data engineering at about 3 to 5 years, which the backgrounds data shows companies explicitly welcome.
How to act on this without reading every posting yourself
One practical takeaway from the data is that forward deployed engineering hiring is fragmented. It is spread across 160 different companies, from horizontal platforms to defense contractors to vertical-AI startups, with no single dominant employer. Finding the roles where your particular mix of stack, vertical, and travel tolerance actually fits is most of the work.
That is the gap MORT is built to close. MORT is an AI job-matching platform for people who want to apply to more relevant jobs in less time. It scans thousands of company career pages, scores every job 0-100% for compatibility with your skills and experience, and generates a tailored resume (a CV, if you are in the UK) for each application. For a role this fragmented and this specific about its requirements, a fit score does the triage so you spend your time on the postings you can actually clear. If you would rather do the tailoring by hand, our guide on how to match your resume to a job description walks through the same idea step by step.
Frequently asked questions
What skills do forward deployed engineers need?
Production software engineering ability, Python first (71% of postings), plus LLM application skills. Two thirds of postings (67%) are about building LLM solutions for customers, with agents (59 mentions), RAG (35), and prompt engineering the working vocabulary, and daily use of agentic coding tools like Claude Code, Cursor, or Codex is now a named requirement. On the human side, communication (76%) and ownership (69%) top every other soft skill. Figures are from MORT Role Insights, July 2026, across 160 postings.
Do you need a degree to become a forward deployed engineer?
Usually not. In MORT Role Insights July 2026 analysis of 160 forward deployed engineer postings, 72% mentioned no education requirement at all. Degree requirements clustered at the extremes, junior pipelines and deep-specialization principal roles. For most postings, production experience substitutes for a degree.
How much do forward deployed engineers get paid?
Of the 59 postings that state a yearly USD range, the median advertised band was $152,000 to $225,000, with a midpoint of $190,000 and top-of-market roles reaching $440,000 (xAI, $180,000 to $440,000). Only about 37% of postings advertise salary at all and the sample skews US, so treat these as US-market bands. Figures are from MORT Role Insights, July 2026.
How much experience do you need to be a forward deployed engineer?
Where a minimum is stated (106 of the 160 postings), the median is 5 years, and 56% of postings are senior or staff-plus versus only 7.5% junior. The realistic entry paths are junior pipelines at large platforms, or lateral moves from solutions engineering, consulting, or data engineering at about 3 to 5 years. Figures are from MORT Role Insights, July 2026.
How much do forward deployed engineers travel?
60% of postings mention travel and 55% involve onsite embedding at customers, though 40% do not mention travel at all. Demands range widely, from about 25 to 30% travel (Sigma, DigitalOcean) to 80 to 100% embedding in classified environments (Yurtsai) and multi-month relocations (Sandtech asks for 12 to 24 months in Tanzania). Figures are from MORT Role Insights, July 2026, across 160 postings.
Find the forward deployed roles you can actually clear
MORT scores every job 0-100% for how well it fits your skills and experience, then generates a tailored resume for each application. For a role spread across 160 very different companies, that is how you spend your time on the postings that fit.
About the data
This analysis is MORT Role Insights for forward deployed engineer, July 2026. It covers 160 live job postings from 160 distinct companies, sampling the newest posting per company (so a heavy poster like Databricks, with 76 listings, counts once). We read and structured each posting individually. Because the sample is one posting per company, it reads the market rather than any single company's hiring volume. Sub-sample figures keep their qualifier: years of experience is stated in 106 of the 160 postings, and salary reflects only the 59 postings that advertise a yearly USD range, a sample that skews US and larger companies. This is part of a set of MORT Role Insights reports; we also analyzed solutions engineers, supply chain and procurement roles, and registered nurses over the same period.