Parses Anything

Formal Grammars For COBOL, SQL, JCL, And 40+ More.

Rule-Driven

Composable Rewrite Rules, Readable And Debuggable.

Deterministic

Same Source in, 100% repeatable results out.

Auditable

Every Transformation Traceable And Reproducible.

What is DMS

DMS® is a mature software transformation platform that has been successfully used for more than 30 years to analyze, understand, and transform complex software systems. Over that time, it has processed more than 500 million lines of source code across 40+ programming languages, making it one of the industry's most proven platforms for large-scale software modernization. Organizations have relied on DMS to modernize mission-critical applications where accuracy, scalability, and preservation of business logic are essential.

While DMS is widely recognized for deterministic code translation, its capabilities extend far beyond language conversion. The platform has been used for code analysis, application refactoring, architecture modernization, business rule extraction, and the development of custom software engineering tools tailored to unique customer requirements.

Whether the objective is translating COBOL to Java, refactoring legacy applications, building custom analyzers, or creating specialized transformation tools, DMS provides a deterministic, repeatable, and scalable foundation for software engineering. Decades of successful deployments across a wide variety of programming languages and modernization projects demonstrate the flexibility and reliability of the platform.

DMS is not a code generator. Not an LLM. It is a deterministic transformation engine designed for zero-defect modernization at scale.

How DMS Works

DMS is engineered specifically for industrial-scale code analysis and transformation. DMS uses classic compiler components like parsing, control-flow analysis, and data-flow analysis — but what makes DMS unique is its symbolic reasoning, pattern matching, and composable rewrite rules that enable deep, large-scale, context-aware code transformations.

Unlike probabilistic AI systems, DMS operates through formal program analysis and deterministic transformation logic.

Collect and Parse Source Code

Legacy repositories are often fragmented, inconsistent, and partially broken.

DMS:

  • Collects and parses all source components targeted for transformation
  • Identifies and reports anomalies such as:
    • Code that no longer compiles
    • Obsolete or dead code
    • Missing copybooks
    • Structural inconsistencies
  • Parses not only COBOL, but also related technologies that must be transformed — including embedded SQL and database access layers

This creates a complete and analyzable model of the system before transformation begins.

DMS VERSUS AI

Deterministic transformation vs probabilistic generation.

Both can produce modern Java, C#, or C++. They differ in how correctness is established.

Feature
Semantic Designs (DMS)
AI-assisted modernization
Translation approach
The ProblemDeterministic transformation using formal rewrite rules
How DMS handles itProbabilistic Gen AI using fine-tuned LLMs
Logic handling
The ProblemRestructures control flow; GO TO nests rewritten into structured code
How DMS handles itTranslates logic as written; refactoring requires prompts
Semantic equivalence
The ProblemProven by design, embedded in the rules
How DMS handles itValidated after conversion through testing
Output consistency
The ProblemRepeatable; identical output every run
How DMS handles itMay vary between runs
Scale
The ProblemDesigned for millions of lines
How DMS handles itBest suited to smaller or incremental efforts
Risk profile
The ProblemLow and predictable
How DMS handles itDepends on testing rigor and review depth

When DMS is the right choice.

Choose DMS when

  • Zero-defect migration of massive, complex systems is the priority.
  • Manual line-by-line review is impractical.
  • You need semantic preservation at scale.
  • Budget and schedule predictability are critical.

AI-assisted fits when

  • You want to upskill developers unfamiliar with COBOL.
  • You plan to redesign business processes significantly.
  • You prefer an incremental, agile approach.
  • You have time for iterative refinement.

Coverage: 40+ source languages and 15 COBOL dialects, mixed-language estates, and mainframe encodings such as EBCDIC.

COBOL PL/I JOVIAL HLASM Java C C++ C# Fortran SQL Natural/ADABAS Verilog and VHDL
Proof

Proven Across Mission-Critical Software Modernization Projects

For more than 30 years, organizations have trusted DMS® to solve some of the most challenging software engineering problems. From defense and aerospace to banking, manufacturing, government, and financial services, DMS has been successfully deployed on projects involving code analysis, automated translation, application refactoring, architecture recovery, clone detection, and custom software engineering tools. Every engagement is built on the same deterministic compiler technology that delivers predictable, repeatable results at enterprise scale.

DMS has processed applications ranging from hundreds of thousands to 200+ million lines of code across 40+ programming languages. Whether modernizing legacy COBOL, translating millions of lines of C++, or building specialized analysis tools, DMS consistently delivers automation that would be impractical through manual effort alone.

Boeing

Boeing

Automated restructuring of 5+ million lines of C++ avionics mission software into a CORBA-compatible architecture for unmanned aerial vehicles.

Northrop Grumman

Northrop Grumman

Automated translation of 1.2 million lines of JOVIAL mission software to C, extending the life of critical B-2 flight software.

Financial Services

Financial Services

Automated conversion of 500,000+ lines of COBOL to C# in less than six months while preserving like-for-like functionality.

Tableau

Tableau

Refactored 4+ million lines of C++ by automatically restructuring APIs and eliminating God Classes.

Australia and New Zealand Bank

Australia and New Zealand Bank (ANZ)

Delivered custom architecture analysis tools for 15+ million lines of Hogan COBOL core banking software.

U.S. Social Security Administration

U.S. Social Security Administration

Developed analysis tools for one of the world's largest legacy environments, spanning 200+ million lines of COBOL, Assembler, SQL, JCL, and Java.

Dow Chemical

Dow Chemical

Built automated migration tools that have modernized 300+ factory control systems, preserving decades of embedded process knowledge while enabling migration to industrial controllers.

Automatic Data Processing

Automatic Data Processing (ADP)

Automated translation and refactoring of HLASM to maintainable C, reconstructed control flow, and optimization.

Sallion Energy

Sallion Energy

Used CloneDR® to identify and refactor duplicate code across multiple large C++ applications, significantly reducing maintenance costs.

Large Global Bank

Large Global Bank

Applied CloneDR® to analyze 10+ million lines of Java and C# source code to identify software plagiarism across hundreds of systems in just two weeks.

These projects demonstrate that DMS is far more than a code translator. It is a comprehensive software transformation platform that enables organizations to understand, modernize, refactor, and evolve complex applications with deterministic precision while preserving the business logic that makes these systems valuable.

30 Source Transformation
40 Source Languages
500 Lines Analyzed
100 Project Success Rate

See what DMS says about your codebase.

1
Free codebase assessment

Send your inventory under NDA. Complexity report, automation estimate, and rough timeline in 5 business days. No fee.

2
Pilot translation

Pick 50,000 lines. We translate them to your target in 2 weeks, shown side by side with the source.

3
Talk to our engineers

Not a salesperson. The engineers who configured DMS for the B-2 bomber and for Dow Chemical.

FAQ

The questions our engineers get
asked the most

General questions

Yes. Same source in, byte-identical output out, every run. The rule library is auditable.

Neither. It is symbolic AI and enhanced compiler technology. Gen AI sits adjacent for documentation and test generation, never for translation decisions.

40+ languages and 15 COBOL dialects, plus mixed-language estates and encodings like EBCDIC. Custom front-ends are added under contract.

Yes. It deploys on your Linux or Windows servers; source, output, and results stay inside your network.

Yes, outright on completion. No runtime royalties, no platform lock-in.

Dont let your last COBOL developer retire before your migration is planned.

COBOL headcount is falling 8% a year. Contractor rates are climbing 12 to 18% annually. Start with a free assessment. Decide afterward.