ConchoBlog
Field notes
Field notes on legacy modernization, code knowledge graphs, AI context, and the hidden system intelligence inside critical software.
- May 2026
Why AI Struggles With Legacy Codebases
AI coding tools can help modernize legacy systems, but only after the codebase has been turned into a structured, queryable map.
- April 2026
The Knowledge Cliff: What Happens When Your COBOL Experts Retire
When legacy experts retire, companies lose more than programming skill. They lose the operating knowledge hidden inside critical systems.
- March 2026
A Private Equity Guide to Technical Diligence on Legacy Software
PE firms inherit more than product and revenue. They inherit architecture, hidden technical debt, and modernization risk.
- February 2026
Modernizing Patient Portals Without Breaking Critical Workflows
Healthcare portals sit on top of fragile workflows, sensitive data, and old integrations. Modernization has to start with a system map.
- January 2026
What Is a Code Knowledge Graph?
A code knowledge graph turns files, functions, workflows, business rules, and dependencies into queryable system intelligence.
- December 2025
MCP for Legacy Systems: Giving AI Agents Safe Access to System Context
MCP gives AI agents controlled access to structured legacy system context, so they can work from a verified map instead of raw files.
- November 2025
The Hidden Cost of “Just Rewrite It”
Rewrites fail when teams underestimate hidden business logic, integrations, and dependencies. Modernization should start with a map.
- October 2025
How to Find Business Rules Hidden Inside Legacy Code
Business rules often live inside legacy conditionals, jobs, integrations, and exceptions. Recovering them is essential before modernization.
- September 2025
Security Discovery in Old Codebases: What AI Alone Misses
Old systems hide security risk in relationships: auth paths, data flows, privileged jobs, dependencies, and integrations.
- August 2025
From Codebase to Executive Briefing: Translating Technical Debt Into Decisions
Executives need more than vague technical debt warnings. They need codebase complexity translated into risk, sequencing, and investment decisions.