Executive summary
A Fortune-scale global retailer needed to understand its 20+ year-old Merchandising Management System, 7M+ lines of entangled and largely undocumented code, before it could commit to any modernization path. The company had been here before. A previous assessment had cost $1M, taken 9 months, and delivered a static document that was outdated before the ink dried.
With Gallop Discover, the retailer's own team reached a deeper, verifiable understanding of the entire system in 3 weeks. It went further than any static assessment could: Discover's planning agents used that understanding to design and compare target architectures for Google Cloud, priced and machine-verifiable. The client walked away with the confidence to make a modernization decision and a living blueprint to execute against.
The challenge: confidence, not just documentation
Before committing a dollar to a rebuild, leadership had to trust what the system actually did. Three problems stood in the way:
- A system nobody fully understood: Two decades of business logic, pricing, inventory, and merchandising rules lived in thousands of undocumented, interdependent stored procedures and cross-module code paths. The engineers who wrote the earliest layers were long gone.
- A failed model for understanding it: The previous attempt at clarity was a conventional consulting assessment: 9 months, $1M, and an army of analysts running interviews and manual code surveys. The deliverable was a static report that could not be queried or verified, and it began drifting from reality the day it arrived.
- A decision no one could make blind: Modernizing a 7M+ line system that runs the merchandising operations of a $5B+ business is not a leap taken on faith. Leadership needed to know what the system does, what depends on what, where the risks are, and what the destination should look like.
The approach: Gallop Discover
Rather than commissioning another manual study, the client ran Gallop Discover against the full estate, in two stages:
- Deep comprehension of 7M+ lines: Discover's comprehension agents parsed 20 years of Java, PowerBuilder, and deeply entangled PL/SQL, resolving cross-module dependencies no human survey could reach. The full codebase and the company's technical documentation were compiled into a knowledge graph: a queryable model capturing not just the system's structure but its business meaning. The retailer's own analysts explored it directly, asking questions in plain language, following dependencies, and surfacing flows the organization had lost visibility into.
- From understanding to destination: Where a traditional assessment stops at what you have, Discover kept going. Its planning agents traversed the knowledge graph to generate machine-verifiable target architectures for Google Cloud, applying domain-driven decomposition to design an event-driven, loosely coupled set of microservices out of the monolith. Architecture options were compared and priced on the platform, and every design decision was reviewed and locked by the client's internal team.
Business impact: from a leap of faith to an engineering call
The engagement gave leadership evidence where it previously had guesswork:
- 3 weeks instead of 9 months: The full deep assessment plus destination architecture was delivered in under a month, more than a 12x acceleration over the retailer's previous assessment cycle.
- A fraction of the cost: The prior engagement consumed $1M for a static report. Discover delivered a deeper result, with a living, queryable model the team still uses daily.
- Confidence to proceed: For the first time, leadership could see, with evidence, what the system does, what is fragile, and what the modern version looks like.
- A reusable blueprint: The knowledge graph and locked architecture blueprints became the foundation for the rebuild itself. Nothing from the assessment phase was thrown away or re-discovered.
- An agent-ready baseline: The target Google Cloud architecture was designed agent-ready from day one, with clean service boundaries, organized data, and infrastructure built for the pace agents operate at.
“We're highly impressed by Gallop Intelligence. In a short window it surfaced insights of remarkable depth, discovering flows we had no visibility into, and the chat interface made those insights genuinely easy to explore.”
C-suite Executive, Fortune-scale Global Retailer