🚀 ASAP CLI - AWS Migration Assistant

ASAP (AWS SQL Analytics Platform) is an AI-powered CLI tool designed to accelerate SQL and ETL migrations to AWS. Built by the AWS ProServ Data & Analytics Latam Team, it provides expert-level migration assistance with interactive guidance and validation.

🎯 Overview

ASAP CLI leverages advanced AI agents to:

• Migrate SQL/ETL from legacy platforms (Oracle, Teradata, SQL Server) to AWS services
• Optimize queries for performance and cost efficiency
• Assess migration complexity and risks
• Generate implementation roadmaps with step-by-step guidance
• Suggest AWS-native architectures using Glue, EMR, Redshift, and Athena
• Validate data quality and provide testing strategies

✨ Key Features

🔄 SQL/ETL Migration

• Multi-platform support: Oracle, Teradata, SQL Server → AWS
• Target platforms: Amazon Redshift, Spark SQL, AWS Glue
• Interactive migration process with feedback loops
• Semantic preservation of business logic and data types

⚡ Query & Transformation Optimization

• Performance tuning recommendations
• Cost optimization strategies
• AWS-specific optimizations (DISTKEY, SORTKEY, partitioning)

🎯 Expert Migration Assessment

• Complexity and risk analysis
• Compatibility issue identification
• Business rules validation
• Migration readiness evaluation

📋 Implementation Planning

• Detailed migration roadmaps
• Step-by-step execution plans
• Field mapping documentation
• Test case generation

🏗️ AWS Architecture Guidance

• Service recommendations (Glue, EMR, Redshift, Athena)
• Best practices implementation
• Performance optimization
• Security considerations

🛠️ Debug & Validation Tools

• SQL debugging and correction
• Data quality validation
• Test result comparison
• Automated fixing with explanations

🚀 Quick Start

Prerequisites

• Python 3.10+
• AWS CLI configured with appropriate permissions
• AWS Bedrock access (Claude models)
• Git for version control

Installation

1. Clone the repository:

   git clone git@ssh.gitlab.aws.dev:jorsie/asap-cli.git
   cd asap-cli
   

2. Install dependencies:

   pip install -r requirements.txt
   

3. Install the CLI tool:

   pip install -e .
   

4. Set up aws account:

   1. Go to your aws account.
   2. Create user with access granted to aws bedrock.
   3. Configure your acces key and secret acces key for aws cli.
   4. Activate your aws cli profile.
   5. Go to aws bedrock and ensure access to Claude 3.7 Sonnet and Claude Sonnet 4

Altenative Installation

1. pip install asap-cli
   

2. Set up aws account:

   1. Go to your aws account.
   2. Create user with access granted to aws bedrock.
   3. Configure your acces key and secret acces key for aws cli.
   4. Activate your aws cli profile.
   5. Go to aws bedrock and ensure access to Claude 3.7 Sonnet and Claude Sonnet 4

Usage

Start the interactive migration assistant:

asap agent run

The CLI will guide you through:

1. Pipeline identification - Name your migration pipeline
2. Source platform - Specify your current SQL dialect
3. Target platform - Choose your AWS destination
4. Migration execution - Follow interactive prompts
5. Validation & debugging - Test and refine results

📖 Usage Examples

Example 1: SQL Server to Redshift Migration

asap agent run

├── migration/                 # Migration workspace
│   └── {pipeline_name}/
│       ├── source/           # Source SQL files

1. The first you need to do is open the directory in your terminal where you are gonna work
2. Create the migration directory and inside that create a new folder with the pipeline name
3. Inside the pipeline name folder add new folder call source and inside this folder create the .sql file you want to migrate. (The system accept .sql and .txt files)
4. After that you are ready to make it your firts migration.

Command example

aws-proserv:~$ I need to migrate my sales pipeline from SQL Server to Redshift

# The agent will:
# 1. Ask for pipeline name: "sales"
# 2. Request source SQL files in migration/sales/source/
# 3. Generate assessment and translation
# 4. Provide interactive feedback loop
# 5. Output optimized Redshift SQL

Example 2: Debugging SQL Issues

├── migration/                 # Migration workspace (use this as an example)
│   └── {pipeline_name}/
│       ├── source/           # Source SQL files
│       ├── target_sql/       # Translated SQL output
│       ├── assessment/       # Migration assessments
│       ├── expected_result/  # Test expected results
│       └── query_result/     # Actual test results

For debugging firts you need to extract two main things:

1. Get an expected result of your query, this is the result from your source sql. Save in expected_result folder inside your pipeline name folder.

2. Get the actual result form the target sql query, this is result you obtain after execute the same values with the target sql query. Save in query_result folder Note: The system only recieve files in .csv format.

3. For the expected_result and query_result you have to assign the same name, this name is your test case name.  The agent use that to make the debugging procees. You could have more that one test case, the only rule you have to follow is same name for each expected_result and query_result files and not repeat test case names.

Example 2: Example on command

aws-proserv:~$ Debug my customer_analytics pipeline with test case "revenue_calculation"

# The agent will:
# 1. Compare expected vs actual results
# 2. Identify discrepancies
# 3. Fix the SQL with explanations
# 4. Save corrected version

🏗️ Project Structure

asap-cli/
├── asap/
│   ├── cli.py                 # Main CLI entry point
│   └── agent/
│       ├── main.py            # Agent command registration
│       └── utils/
│           ├── agents/
│           │   ├── config.py              # Agent configuration
│           │   ├── factory_agent.py       # Agent factory
│           │   ├── migration_assistant_agent.py  # Core agent logic
│           │   └── tools/
│           │       ├── translate/         # SQL translation tools
│           │       ├── debug/             # Debugging tools
│           │       └── general/           # Shared utilities
│           └── general/
│               └── ui.py                  # User interface utilities
├── migration/                 # Migration workspace (use this as an example)
│   └── {pipeline_name}/
│       ├── source/           # Source SQL files
│       ├── target_sql/       # Translated SQL output
│       ├── assessment/       # Migration assessments
│       ├── expected_result/  # Test expected results
│       └── query_result/     # Actual test results
├── requirements.txt          # Python dependencies
├── pyproject.toml           # Package configuration
└── README.md               # This file

🔧 Configuration

AWS Bedrock Models

ASAP CLI uses multiple Claude models for different tasks:

• Primary: us.anthropic.claude-sonnet-4-20250514-v1:0
• Fallback: us.anthropic.claude-3-7-sonnet-20250219-v1:0

📚 Documentation

Core Components

• MigrationAssistantAgent: Main conversational agent with tool integration
• TranslateSQL Tool: Handles SQL dialect conversion and optimization
• DebugSQL Tool: Identifies and fixes SQL issues with explanations
• BedrockLLamager: Manages AWS Bedrock model interactions with failover

Migration Process

1. Assessment Phase:

   • Analyzes source SQL complexity
   • Identifies compatibility issues
   • Maps business rules and requirements

2. Translation Phase:

   • Converts SQL to target dialect
   • Applies AWS-specific optimizations
   • Preserves semantic accuracy

3. Validation Phase:

   • Compares expected vs actual results
   • Provides debugging assistance
   • Generates corrected SQL with explanations

🔍 Troubleshooting

Common Issues

"Could not load agent command"

# Ensure Python path includes the project directory
export PYTHONPATH="${PYTHONPATH}:$(pwd)"

AWS Bedrock Access Denied

# Verify AWS credentials and Bedrock access
aws bedrock list-foundation-models --region us-east-1

Migration Files Not Found

# Ensure proper directory structure
mkdir -p migration/{pipeline_name}/{source,target_sql,assessment,expected_result,query_result}

📊 Supported Platforms

Source Platforms

• Oracle (PL/SQL, Oracle SQL)
• Teradata (Teradata SQL)
• Microsoft SQL Server (T-SQL)
• Generic SQL (ANSI SQL)
• Other SQL platforms (ANSI SQL)

Target Platforms

• Amazon Redshift (PostgreSQL-based)
• Spark SQL (Apache Spark)
• AWS Glue (PySpark/Scala)
• Amazon Athena (Presto SQL)

👥 Team

Developed by: AWS ProServ Data & Analytics Latam Team
Author: Jorge Sierra (jorsie@amazon.com)
Version: 0.0.1

🆘 Support

For support and questions:

• Internal AWS Users: Contact the ProServ Data & Analytics Latam Team, jorsie on slack
• Issues: Use the internal GitLab issue tracker

---

🚀 Ready to accelerate your data migration journey with AWS ProServ expertise!