InsurCore Data extracts and converts Paradox database files — including the tables, memo fields, and relationships that most export tools miss — into clean, modern, import-ready data.
The Problem
Paradox databases use a proprietary format that modern tools often can't read without the original runtime or BDE (Borland Database Engine) installed.
Tables are stored as individual .DB files with separate index (.PX), memo (.MB), and validity check (.VAL) files — a missing companion file can make data unreadable.
Field names and data types are inconsistently defined across old Paradox versions, and there is rarely any documentation of the schema.
Memo fields (long text, notes) are stored in separate .MB files and are often missed or corrupted during manual export attempts.
The software that ran on these databases may no longer exist, be licensed, or run on modern operating systems — making it difficult to even open the data.
Data quality is often poor after years of unvalidated entries — duplicates, broken relationships, and inconsistent formats are common.
What We Extract
Paradox databases are one of the most commonly encountered legacy formats in industries like insurance, healthcare, legal, and field service. We have purpose-built tooling and extraction scripts specifically for this format.
The Process
We inspect the source files, database, or export structure to understand what exists and what condition it's in.
We identify tables, fields, relationships, record counts, duplicates, and critical business objects.
We map the source data to the destination system or desired output format, including any custom field logic.
We clean, normalize, restructure, and convert the data using scripts, tools, and custom logic.
We deliver clean, structured, import-ready data files with a validation report and audit trail.
FAQ
No. We use specialized extraction tools and custom scripts that read the Paradox binary format directly — no Borland Database Engine or original software required.
We do a file integrity check first and let you know what's recoverable. Partial recovery is possible in many cases — we document what was extracted and what was lost.
We deliver data in your target format — typically CSV or Excel for import into modern platforms, or SQL insert scripts for database targets. We can also load directly into PostgreSQL or SQLite.
Not usually. We've worked with very old Paradox databases. The older the data, the more likely there are encoding issues (old code pages) — but we handle those as part of the normalization step.
We offer a free initial data review — no commitment required. Send us a sample file or describe your source system and we'll tell you exactly what's possible.