Skip to main content

← Back to Topics

Enterprise and boutique implementers, and the forward-deployed-engineer wave

How the best-funded AI implementers work, what the forward-deployed-engineer model costs to run, and how far down-market it reaches as of October 2026.

Tiers A and B of the market map are the firms that put engineers inside a client's operations to build custom AI. The industry now calls this forward-deployed engineering (FDE), a term taken from Palantir. This page covers who the funded players are, what the model costs, and where it stops.

Tier A and B: the funded implementers

CompanyScaleModel
Turing (San Francisco, 2018)About $225M raised; $2.2B valuation at its March 2025 Series E; about $300M annual recurring revenue (TechCrunch)Data and coding work for AI labs, plus custom AI systems for Fortune 500 clients
Distyl AI (San Francisco, 2022)$175M Series B at $1.8B, Sept 2025; the company says revenue grew 5x in 2024Embedded engineers plus its own agent platform, with results aimed within about a quarter
OpenAI Deployment Company (2026)More than $4B of initial investment, TPG leading; about 150 engineers and specialists through the Tomoro acquisitionEmbedded engineers at large enterprises; an earlier report said OpenAI consulting charged at least $10M per client
Invisible Technologies (New York)A round of about $100M at a valuation above $2B, reported by Bloomberg in Sept/Oct 2025; revenue figures come from aggregators onlyServices plus a platform for enterprises and AI labs
Tribe AI (2019)$3.25M seed (July 2024); an eight-figure revenue run rate claimed by its CEO; a network of 500+ contractorsEmbedded experts in three phases: Map, Build, Activate
Fractional AI (San Francisco, 2024)About 44 people and about $4.8M revenue (LinkedIn and Latka estimates, Nov 2025); about 88 people at acquisition per PlankCustom generative-AI builds; acquired in May 2026 by Anthropic's enterprise-services venture, terms undisclosed
Cassidy (New York)$10M Series A, Sept 2025No-code agent platform plus hands-on implementation; names insurance and professional services among its segments

Three patterns:

  • Revenue at the top is often data work. Turing's scale comes mainly from supplying expert human time to AI labs, with custom implementation as a second line.
  • A services business can grow on little venture money. Tribe AI claims an eight-figure run rate on a $3.25M seed.
  • Prices are rarely public. Among these firms only product-led ones publish a pricing model, and none of the services firms publish project prices.

Tribe AI's homepage reads "We embed. We build. We don't stop until it's in production." Anthropic's venture, which now includes Fractional AI, says "We don't hand off a repo and disappear."

The wave

Palantir built its business on FDEs. In Q2 2026 its US commercial revenue rose 149% year on year to $764M, it reported a record 220 deals of $1M or more in the quarter, and it had 653 US commercial customers (Palantir results via BusinessWire and the earnings call transcript). Its sales motion is a bootcamp in which the customer brings its own data.

Hiring. Forward-deployed engineer job postings rose 729% in the year to April 2026 (Indeed index, via Business Insider and Plank). Other windows give other figures: more than 800% over January to September 2025 (Indeed and the Financial Times, via Pepper Eyes) and 1,165% year on year in Revealera data from November 2025 (Bloomberry, via Plank). On 18 October 2026 Plank counted 1,262 live roles at 555 companies, with a posted median pay band of $160K to $230K.

The labs and platforms are moving in.

  • OpenAI launched its Deployment Company in May 2026.
  • Anthropic announced Ode on 4 May 2026, a services venture of about $1.5B aimed at mid-sized companies such as community banks, mid-sized manufacturers and regional health systems. Plank estimates its clients at $50M to $2B in revenue.
  • Microsoft and AWS announced forward-deployed programmes of $2.5B and $1B respectively (2026, via Plank).
  • Sierra (customer-service agents) and Decagon bill per resolved conversation, with engineers bundled in.

The economics

Two published analyses put the break-even for lab-style FDE work in the hundreds of thousands of dollars per client per year:

  • Pepper Eyes (May 2026, author calculations) finds that long-tail engagements under about $500K a year lose roughly $500K to $700K each, and that work at $0.5M to $3M is marginal.
  • N47 (Sept 2025) says the model fits enterprises paying $1M or more a year and is a poor fit for small business and self-service buyers.

ICONIQ's operator survey reports how software firms package it: 30% bundle FDE support into the subscription, 29% charge a separate services fee and 25% do a hybrid. Examples include $40K to $45K annual contract value with a week of FDE support, and a $500K six-month pilot.

Neither analysis addresses a small studio whose costs are not those of a lab engineer (a fully loaded lab FDE is estimated at $845K a year).

How far down-market it reaches

Labs: self-serve rather than embedded. Anthropic launched Claude for Small Business on 13 May 2026 with 15 workflows; by 15 September it had 43 workflows and 27 more integrations. It has started recruiting small-business-focused consultancies into its Claude Partner Network. OpenAI has announced a ChatGPT small-business programme with webinars and guides. Neither lists embedded engineers for small firms.

Small providers using FDE language. One directory lists 34 providers, 9 of which publish prices; one entry, Brilliantech, lists embedded engineers at $15,000 to $40,000 a month.

ProviderOfferPublished priceClient size
forwarddeploy.coAI for CPA firmsnone listedCPA firms of 15 to 80 people
Law Tech AICoaching and implementation for law firmsnone stated1 to 75 attorneys
aideployed.devFree audit, setup, retainersetup "under $5k", retainer "under $1,000/mo" (ceilings)small and medium businesses
EupraxicOne workflow in 10 business days$5,000, refund guaranteefounder-led B2B software firms
UtsuboPod of two to four for three to six monthsfixed monthly fee, unpublished11 to 500 employees
Phos AI LabsAudit plus embedded engineeraudit from $10,000; embedded from $15,000 a monthUS mid-market, $5M+ revenue
BrilliantechProof of concept, embedded, podsproof of concept from $8,000; $15K to $40K a monthnot stated
A.TeamVetted engineers under the client's brandabout $25,000 a month per senior buildersoftware companies

Few of these publish proof. The only small-firm results found with figures are self-reported by forwarddeploy.co (for example a workpaper review cut from 12 hours to 3 in an 80-person firm).

Companies mentioned

Sources