Legacy modernization, systems integration, and AI reliability

$291K earned across 213 Guru transactions for 17 employers.

I take ownership of difficult business systems - software that is failing in production, too important for a big-bang rewrite, or understood only in pieces across the interface, database, integrations, and operating workflow.

I have been a working programmer since 1979 and a Guru member since 2004. The record includes portfolio and accounting products, cross-platform C/C++, Microsoft Access and Excel automation, SQL Server and T-SQL, payment processing, Salesforce and Sage integrations, web and native mobile applications, complex algorithms, and current AI-assisted engineering.

My work now concentrates on three connected services:

1. Legacy rescue and controlled modernization

I stabilize, repair, and replace Microsoft Access/VBA, SQL Server, desktop, and low-code systems without asking the business to stop while the rewrite happens. The first job is often discovering which source actually runs production, which data is authoritative, and which undocumented behaviors people depend on.

2. API integration and data reconciliation

I connect systems that were never designed to cooperate: payment processors, accounting platforms, CRM systems, operational APIs, databases, and custom applications. A successful request is not treated as proof that the business event completed. Counts, diagnostics, missing-record reports, and replayable checks make quiet failures visible and recoverable.

3. AI reliability, algorithms, and model validation

I translate complex formulas into tested software and build AI-dependent workflows in which architecture, data authority, acceptance, and release remain explicit engineering decisions. Models can widen the search and challenge the work; they do not receive the final say merely because their output sounds confident.

My method is consistent across all three: establish what is actually true, define acceptance before implementation, deliver in controlled slices, independently check important outcomes, and preserve a way back until the new path has earned trust.

The technology changes with the constraint. The responsibility does not. You work directly with the principal engineer who diagnoses the problem, builds the solution, and remains accountable for whether it works in operating reality.

Everyone calls me Andy.

Work Terms

Engagement: I prefer to begin with a bounded problem, the current symptoms, the systems involved, what must keep working, and the decision or deadline the work must support. For uncertain systems, the first milestone may be a diagnostic and evidence package before implementation begins.

Communication: Written requirements, priorities, decisions, and acceptance criteria keep the work clear. Calls are useful for discovery and difficult decisions; important conclusions are documented afterward.

Availability: I work remotely from New Jersey and schedule focused project time based on the agreed scope, milestones, and response requirements. Urgent production work is arranged explicitly rather than promised as unlimited availability.

Payment: Guru SafePay or SafePay Escrow, with milestones matched to inspectable deliverables and acceptance evidence.