SSMA Copilot for SAP ASE: AI-Powered Migration to SQL Server

Microsoft has unveiled SSMA Copilot for SAP ASE (formerly Sybase), an AI-assisted migration tool that addresses one of database modernization's most persistent challenges: converting complex Sybase procedural code to SQL Server-compatible syntax.

The new capability builds on SSMA's existing rule engine, which already handles approximately 70% of standard Sybase-to-SQL Server conversions. However, Sybase's extended T-SQL dialect—with its proprietary syntax variations, non-standard conditional logic, legacy system tables, and database-scoped behaviors—has long frustrated migration teams. These nuances make stored procedures, triggers, and packages particularly difficult to convert automatically.

Why Traditional Rule-Based Conversion Falls Short

Rule-based systems excel at predictable patterns but struggle with Sybase's unique constructs. When encountering proprietary syntax, conditional logic expressed in non-standard forms, system-level commands unsupported in SQL Server, or ambiguous constructs requiring contextual interpretation, traditional converters often fail or produce commented fallback code that requires extensive manual rewriting.

This gap typically forces developers into tedious manual rewrite processes, extending migration timelines and increasing the risk of errors. The 30% of code that rule engines cannot handle often represents the most complex and business-critical components of an application.

How Copilot Changes the Migration Game

SSMA Copilot introduces agentic AI to fill these conversion gaps. Rather than relying solely on static rules, the AI system understands intent, identifies root causes of conversion failures, and generates SQL Server-compatible alternatives with transparent reasoning.

The implementation follows a similar architecture to the recently released Oracle-to-SQL Copilot, bringing consistent AI-assisted migration capabilities across Microsoft's database modernization portfolio.

Authentication Flexibility

SSMA for SAP ASE offers two authentication methods to accommodate different organizational requirements:

• Bring Your Own Key: Connect to your Azure OpenAI resource using your deployment details and API key. This option suits organizations that already manage Azure OpenAI or require strict control over their AI environment.

• Microsoft-Managed Endpoint (Preview): A seamless experience requiring no API key. Users sign in with Microsoft Entra ID, and SSMA handles authentication through a secure browser-based flow.

The Tri-Pane Conversion Experience

When users trigger the "Fix with Copilot" button, SSMA presents a structured interface designed for clarity and trust:

1. Errors to Fix: Displays issues the rule engine could not convert—whether due to unsupported syntax, parse failures, or ambiguous constructs. This helps users quickly identify where the rule engine struggled.

2. Explanation: Provides a detailed, human-readable breakdown explaining why the conversion failed, what the Copilot-generated fix means, and how the logic differs from Sybase to SQL Server. This transparency builds trust in AI-generated code.

3. Code Review Window: Shows a side-by-side diff with SSMA-generated output on the left and Copilot-converted SQL code on the right. Changes are highlighted so users can validate improvements, understand transformations, and decide whether to apply the Copilot output.

Real-World Impact: A Practical Example

Consider a Sybase stored procedure that fails conversion because it uses set switch on drop_system_tables with override, no_info—a Sybase-only command unsupported in SQL Server. The procedure definition also contains create or replace procedure, which is not valid T-SQL syntax.

When the rule engine encounters these issues, it cannot parse the affected block and outputs the original Sybase procedure as a commented fallback. However, when Copilot is invoked:

• It identifies the unsupported keywords
• Suggests correct SQL Server equivalents (e.g., translating create or replace into IF EXISTS ... DROP + CREATE PROCEDURE)
• Generates a complete, runnable T-SQL procedure
• Explains why each fix was made

This allows users to resolve previously conversion-blocking issues instantly, transforming what could be hours of manual work into a guided, high-confidence workflow.

Strategic Implications for Database Modernization

The introduction of AI-assisted conversion represents a significant evolution in database migration strategy. Organizations can now:

• Accelerate migration timelines: Reduce manual fix-up effort and speed up conversion cycles
• Improve code accuracy: Generate syntactically correct, context-aware SQL Server code
• Enhance maintainability: Benefit from clear explanations that improve learning and long-term code understanding
• Scale modernization efforts: Handle complex procedural conversions without depending on scarce Sybase expertise

For financial services, telecom, and large-scale enterprise applications that rely on Sybase for mission-critical workloads, this capability removes a major barrier to cloud modernization. The combination of deterministic rule engine and adaptive AI creates a more complete, scalable, and user-friendly migration experience.

As database modernization continues to accelerate, tools like SSMA Copilot demonstrate how AI can address the nuanced challenges that have historically made certain migrations prohibitively complex or resource-intensive. The result is a more accessible path to SQL Server and Azure SQL for organizations of all sizes running legacy Sybase workloads.

For organizations planning Sybase migrations, the new Copilot experience transforms complex procedural conversions into guided workflows that maintain code integrity while dramatically reducing manual effort. This represents not just a technical improvement, but a strategic enabler for broader cloud adoption and modernization initiatives.