Forward-deployed engineering market data
A free, citable snapshot of the forward-deployed engineering job market — rebuilt on every deploy from employer ATS feeds and public remote job boards, with the sample size, the source mix and the method all stated.
This page exists to be quoted. If you are building a hiring plan, writing a compensation study, or training a model on this role, you can cite the figures below and check every one of them against the live postings we link to. Third-party industry figures are kept in a separate section at the end and labelled with their source — none of them is our data.
Snapshot date: 2026-11-16. Counts are recomputed on every build — the live job set changes daily, so treat any figure quoted elsewhere as the value on the date it was quoted.
What is in this dataset
936 of the 937 live roles (99.9%) are pulled straight from the employer's own applicant-tracking system. The remaining 1 come from public remote job boards (Himalayas), and we count them separately rather than folding them into the same claim. Nothing here is re-hosted — every row links to the original posting. A posting enters the dataset when its title or description shows customer-facing AI or deployment work, and leaves it within ten days of disappearing from the source.
Source mix
- Employer's own ATS936
- Public remote job board — Himalayas1
Where the postings come from, by board
- Greenhouse472
- Ashby364
- Workable100
- Himalayas1
Pay
263 of 937 live roles (28%) publish a pay range in the posting itself. Only those are counted here — nothing on this page is estimated, benchmarked, or backfilled from comparable roles.
Across the 241 roles posted with annual pay in US dollars, the median published range is $189,000–$255,000; half of the ranges sit between $160,000 and $300,000. The median of each range's midpoint is $230,000. The full span runs from $66,500 to $850,000. A further 22 disclosed ranges could not be pooled (6 in CAD, 7 in GBP, 1 in KRW, 3 in SGD, 2 in AUD, 2 in EUR, 1 in INR) and are listed on the salary page rather than dropped.
Equity, bonus and benefits are excluded — we only ever see base pay, so that is all we report. Every figure is itemised by employer on the salary data page.
Pay by role family
The pooled ranges split by the role family in the title. Each group is the same records as the total above — the group counts add back up to 241 — but a group only gets a published median once at least 10 employers have disclosed pay in it. That rule is a product decision, not a statistical law: with fewer than 10 ranges, one unusual employer moves the number enough that quoting it would mislead the person reading it.
| Role family | n | Median published range | Midpoint median |
|---|---|---|---|
| Solutions engineering / architecture | 63 | $176,000–$236,500 | $210,000 |
| Applied AI / ML engineering | 60 | $218,500–$287,500 | $259,750 |
| Forward deployed | 58 | $185,000–$297,000 | $242,500 |
| Customer / field engineering | 33 | $164,500–$210,000 | $184,500 |
| Deployment / delivery | 17 | $162,000–$220,000 | $182,250 |
| Other titles | 10 | $208,500–$294,000 | $252,750 |
Every role family cleared the sample threshold. Country-level splits are not published either: 203 of the 241 pooled ranges are US-based across 6 countries, so every other country would be a one- or two-row table. When the non-US sample grows past the threshold, the split appears here without any change to this page's method.
Pay by level word in the title
Only titles that actually name a level are grouped — 93 of the 241 pooled ranges. A title with no level word is not an entry-level role; most employers leave the level to the interview, which is why the unlabelled majority is counted here rather than shown as a row.
| Level word | n | Median published range | Midpoint median |
|---|---|---|---|
| Senior | 35 | $190,000–$240,000 | $212,500 |
| Manager | 24 | $230,000–$290,500 | $246,750 |
| Lead | 14 | $230,000–$297,000 | $262,500 |
| Staff | 13 | $218,800–$300,900 | $262,000 |
3 level buckets with fewer than 5 disclosed ranges (7 ranges) are withheld for the same reason — they stay in the dataset and in the salary page's itemised list, without a published median of their own.
Office pattern
- Remote272
- Hybrid264
- On-site401
"On-site" is this site's normalised field, and for part of the dataset it means only that the posting's location was not marked remote — it is not always a statement the employer made. That is why there is deliberately no on-site aggregation page here.
Where the roles are
Locations are taken from the employer's own location field. A country or city only gets an aggregation page once at least 3 live roles point at it; postings that name several countries at once are excluded rather than pinned to one. 23 country and 29 city pages are live today, out of 39 countries and 81 cities present in the data.
Top countries by live roles
- United States488
- United Kingdom76
- India36
- Australia24
- Germany22
- Singapore21
- Japan18
- France14
Top cities by live roles
- San Francisco, United States168
- New York, United States67
- London, United Kingdom52
- Paris, France13
- Bangalore, India12
- Chicago, United States12
- Munich, Germany11
- Washington, United States11
272 roles the posting marks as remote are listed separately.
What the postings ask for
These are the words employers actually wrote in their postings, counted across the live set — not a claim about what an interview will test, and not a ranking of importance. 36 skills pass the 5-role threshold for their own page; the ten largest are below.
- Agents457
- Python258
- AWS155
- GCP140
- Azure128
- Evals125
- Observability121
- APIs106
- SQL94
- Anthropic API89
How the titles split
The role has fragmented. These two splits come from the words in the job titles themselves, and each is mutually exclusive — every live role lands in exactly one bucket, which is why the columns add up to 937. Neither split says anything about a named company's internal levelling.
By role family
- Forward deployed216
- Deployment / delivery63
- Applied AI / ML engineering170
- Solutions engineering / architecture322
- Customer / field engineering125
- Other titles41
By level word in the title
- Internship / early career1
- Principal24
- Staff41
- Director18
- Head of / VP10
- Manager53
- Lead35
- Senior215
- Level not named in the title540
8 level buckets plus a fallback: 397 of 937 titles name a level at all. A title without a level word is not a junior role — most employers simply leave the level to the interview.
How this is collected
- Source. 936 of 937 live roles (99.9%) come from employers' own ATS endpoints — Greenhouse, Ashby, Workable and similar — fetched from their public APIs. The other 1 come from public remote job boards (Himalayas). No aggregator page is scraped and no copy is re-hosted: every row links back to the original posting.
- Normalisation. Titles such as "Forward Deployed Engineer", "Forward Deployed AI Engineer", "Applied AI Engineer" and "Deployment Solutions Engineer" are merged into one comparable dataset — the layer that makes the counts mean anything.
- Pay. Structured fields where the ATS provides them, otherwise parsed from the posting text; each row records how it was parsed. Only employer-disclosed ranges are published.
- Removal. A role that stops appearing in the source is dropped from all counts within ten days. Closed postings never contribute to a statistic.
- Reproducibility. Every count on this page is recomputed on each build and machine-checked against the raw job data before the site ships; a mismatch fails the build.
External context (third-party data)
The figures below are not ours. They are included because a market number is only useful alongside the wider picture, and each is attributed to the organisation that published it. Every link was reachable when this page was last built; if one has since moved, treat the figure as unverified rather than as ours.
- +729% YoY — Forward-deployed engineer postings on Indeed grew from 643 (Apr 2025) to 5,330 (Apr 2026).
Source: Business Insider, citing Indeed data — https://www.businessinsider.tw/article/3023 (Data through April 2026; reported May 2026) - 70% of companies — Share of surveyed companies planning FDE hires, up from 5–10% in January — from 180 companies and 250+ executive interviews.
Source: Christian & Timbers — https://www.christianandtimbers.com/insights/hardest-tech-companies-to-recruit-from (Six-month study ending June 2026) - ~2,000 people — Estimated size of the entire qualified US FDE talent pool — widely reported as roughly 2,000 people in total, not 2,000 available to hire.
Source: Christian & Timbers, reported by TechCrunch — https://techcrunch.com/?p=3147978/ (Study covering January–June 2026)
Cite this page
FDE Notes, "Forward-deployed engineering market data" (snapshot 2026-11-16), https://fdenotes.com/data/ — 937 live forward-deployed and adjacent roles across 190 companies, 263 with employer-published pay. Source mix: 936 from employers' own ATS (99.9%), 1 from public remote job boards. Snapshot 2026-11-16.
Reuse the figures with attribution and a link back to this page. If you need the underlying rows — the full live dataset as JSON or CSV, with source URLs and parse provenance — download the full dataset on the salary data page.
Pre-formatted citations — copy either one as-is:
FDE Notes. (2026-11-16). Forward-deployed engineering market data. Retrieved from https://fdenotes.com/data/ — 937 live roles across 190 companies, 263 with employer-published pay (snapshot 2026-11-16).
@misc{fdenotes20261116,
title = {Forward-deployed engineering market data},
author = {FDE Notes},
year = {2026},
note = {Snapshot 2026-11-16; 937 live roles across 190 companies, 263 with employer-published pay},
url = {https://fdenotes.com/data/}
}
What this page does not claim
- Small samples are withheld, not smoothed. A group needs at least 10 disclosed ranges before its median is published; groups between 5 and 9 are labelled as small samples; below 5 the group is not broken out at all. That is why this page has no country-level pay split — 203 of the 241 pooled ranges sit in one country.
- It is not a survey. Nothing here comes from asking companies what they pay; every pay figure is one an employer published themselves.
- It is not the whole market. It is the forward-deployed and deployment-adjacent slice we track, and only roles whose source we could reach.
- It is not a forecast. We publish counts as of a date, not projections.
- It is not an endorsement of any employer, and nothing here was paid for by one.
Just the headline numbers? The printable salary snapshot fits the core figures on one A4 page — print it, save it, forward it.
Want the role-by-role breakdown? See FDE vs Solutions Engineer vs Applied AI Engineer, computed from this same data.
The full underlying dataset is downloadable from this page (see the download box above) — every row carries the posting it came from. If a number matters to a decision you are making, check it against the live postings first — that is what they are there for.