AI Is Commoditizing Code. In Travel Tech, That Doesn't Kill Outsourcing — It Relocates the Value

Fred Bean, founder and CEO of HotelPORT, argues AI is restructuring software outsourcing by commoditizing the thing it used to sell: developer hours. The traditional pitch was cheaper labor at scale; as AI coding tools make writing code faster, that arbitrage erodes, because, in Bean’s words, when “three people can produce that outcome,” customers won’t “pay for ten.” His argument is that the industry is shifting from hourly billing toward outcome-based engagement, and that value is migrating from the ability to write code to the ability to know “where the code belongs” — deep contextual knowledge of legacy systems and integrations. He uses travel technology as his specific example: the industry runs on decades-old infrastructure — CRS, PMS, and GDS standards that predate cloud computing — meaning even the best AI coding agent still has to interface with genuinely old, idiosyncratic systems it wasn’t trained to reason about cleanly. His proposed formula for surviving vendors: domain expertise times integration knowledge times AI leverage, not raw coding throughput.

That collision between AI enthusiasm and legacy infrastructure isn’t unique to travel tech. Forrester analyst Mark Moccia made the general case two weeks earlier, warning in a widely-applicable piece that enterprises everywhere are pursuing AI initiatives while quietly still running critical operations on outdated platforms like Access databases — what he calls “Voldemort systems,” the feared legacy platforms nobody wants to touch but that quietly block customer value. His line, “AI won’t save you from your pile of Access databases,” applies just as directly to a CRS built decades ago as to a departmental spreadsheet system: modernizing the legacy layer is a prerequisite for AI value, not a parallel workstream AI can route around.

For hospitality technology buyers evaluating build-versus-outsource decisions, the practical read is that Bean’s formula and Moccia’s warning point the same direction: the vendors and internal teams worth paying are the ones who actually know where the old code — and the old data — lives, not the ones simply offering AI-accelerated throughput.