Find the migration hotspots
Read source code, metadata and application coupling together. The output is a prioritised risk picture, explicit questions and a focused PoC scope.
AssessmentThe real effort is usually buried in packages, PL/SQL, database links, Scheduler jobs, dynamic SQL and application coupling. I make those dependencies visible, validate the hard parts against representative code, and carry the migration through to production cutover.
Assessment · Proof of Concept · Migration · AWS RDS/Aurora
Identify packages, PL/SQL, database links, jobs, dynamic SQL and the Oracle-specific dependencies that drive migration effort.
Implement one representative business flow on PostgreSQL or Aurora and validate code, deployment, performance and failure behaviour.
Bring schema, database logic, Flyway deployments, data movement, testing and cutover together in controlled waves.
A converted schema is not a migrated application. The real question is which behaviour is coupled to Oracle today—and how that behaviour will be proven on the target.
Packages, procedures, functions and triggers may combine API behaviour, state, privileges and business rules.
Database links, Scheduler jobs, materialized views, UTL_FILE and Advanced Queuing need to be understood by purpose, not merely translated by syntax.
Deployment, reconciliation, restart, performance, monitoring and cutover must work on the target—not just the SQL syntax.
Read source code, metadata and application coupling together. The output is a prioritised risk picture, explicit questions and a focused PoC scope.
AssessmentUse a representative business flow, not a demo table. This exposes conversion patterns, extensions, performance and deployment issues early.
Proof of conceptImplement schema and code changes, Flyway, data movement, regression testing, migration waves and rollback through production cutover.
Migration deliveryPackages can combine APIs, state and business logic. Separate those responsibilities before deciding how each part should live in PostgreSQL.
Read the articleJobs are part of the operating model. Capture schedules, dependencies and failure behaviour before choosing the target scheduler.
Read the articleA database link is more than a connection. Classify reads, writes, consistency and failure behaviour before choosing the replacement.
Read the articleMy work includes Oracle lift-and-shift into AWS, PostgreSQL/Aurora environments on AWS, Flyway-based database deployment and the definition of technical API interfaces.
For Oracle → PostgreSQL I focus on the areas where automated conversion alone is not enough: PL/SQL, packages, dependencies, interfaces, deployment, testing and operational behaviour.
Assessment, proof of concept and implementation stay with the same technical lead, so source-system findings feed directly into target architecture, migration units and cutover planning.
For an initial discussion, Oracle version, approximate PL/SQL footprint, intended target and the main unresolved question are enough. No credentials or database content are required.