Every major vendor that scaled forward deployed engineering, or FDE, this year reached for the same lever first: money. Amazon Web Services committed a billion dollars from its own balance sheet in June to embed engineers with customers (AWS, 2026). OpenAI and Anthropic went outside for it, each backing a deployment vehicle with outside equity from private capital rather than funding the work off their own books, a split covered here in July. Infor is trying a fourth lever. Its AI Adoption Hub scales FDE without new capital or new headcount, using AI to do work its own forward-deployed engineers used to do by hand (Infor, 2026).

Three Vendors Chose Capital. Infor Is Choosing Software.

This publication named FDE as enterprise AI's real differentiator back in February, months before the current wave of vendor investment: engineers who set up software to match a customer's specific workflow beat an off-the-shelf product tuned for a generic majority of use cases. The question since then has been how vendors fund enough of those engineers to meet demand. AWS answered with balance-sheet cash. OpenAI and Anthropic answered by selling equity in dedicated deployment companies. Salesforce answered with a partner network, tying outside systems integrators' revenue to Agentforce deployments and structuring the implementation risk into that partner network rather than its own product team, as this publication noted in May.

Three different vendors, one identical lever: capital.

AI Now Drafts the Blueprints Engineers Used to Write by Hand

Infor's AI Adoption Hub runs four stages, and AI handles the first pass at each. Prototype: AI drafts a solution blueprint from industry requirements, cutting scoping time from weeks to days. Refine: AI tests that blueprint against real workflow data before a line of code ships. Deploy: automated pipelines push the validated build into production with governance intact. Evolve: once live, the Hub keeps managing updates and new capability requests, so nothing routes back through a full scoping cycle (Infor, 2026).

Team Air Distributing shows what that pace looks like in practice: three agents built in under two weeks. A separate industrial manufacturing customer got a complete agent spanning the AI and application layers in that same window, a case study this publication examined in July for what it reveals about Infor's account-deepening sales pattern (Infor, 2026).

Rick Rider, Infor's SVP of AI Innovation, put it plainly: "Deployment is not the finish line" (Infor, 2026).

Four Models, One Unanswered Question

AWS's engineers come from its own frontier AI teams and report to AWS. OpenAI and Anthropic's deployment vehicles answer to outside investors who now hold a stake in how those engagements run. Salesforce's engineers work for systems integrators paid to deepen the deployment. Infor's engineers stay Infor's engineers, freed from prototyping and deployment work to spend more time advising customers on strategy, with the AI Adoption Hub doing what capital did for the other three (Infor, 2026).

Each model answers the funding question differently. The open question is what happens after: once a vendor's FDE capacity scales, whether through cash, equity, partners, or AI tooling, how much of that capability transfers to the buyer's own organization, and on what timeline. AWS states a goal of leaving customers self-sufficient with their own AI engineering capability. Infor says customers will get direct access to the Hub in the coming months (AWS, 2026; Infor, 2026).

About Shashi Bellamkonda

Loading bio…

LinkedIn Profile

Get new posts by email:

Subscribe

Powered by