Cloud

Insights

Introduction

The technological landscape of any enterprise includes public cloud, on-premises systems, and hybrid environments. Maintaining accurate documentation is challenging due to frequent changes, leading to documentation drift. SystemViz addresses these challenges by making documentation dynamic and automated, improving maintainability and compliance.

Solution Overview

The conventional documentation process follows a well-known yet imperfect pattern: architects develop detailed diagrams during the planning phase, but deviations occur during implementation due to practical constraints. Over time, operational changes further alter the environment, yet updates to documentation are often overlooked due to time pressures and competing priorities. This cycle results in documentation drift, where the recorded state diverges from the actual deployed technology stack.

The SystemViz solution developed by Infosys transforms documentation into a dynamic, living asset for any technology environment. By using the generative AI capabilities in Amazon, including Bedrock and OpenSearch Serverless, the solution automatically analyzes and visualizes configurations from legacy systems, Infrastructure as Code (IaC), container orchestrations, and traditional IT setups.

From data centers to cloud-native architectures, this solution empowers organizations to operate with clarity and agility, fostering innovation while enhancing governance across their entire technology estate. It eliminates manual efforts in documentation management and ensures alignment between deployment reality and its representation in documents by generating deployment diagrams and comprehensive documentation from the deployment artifacts.

Fig 1

Automated System Understanding

Intelligently parsing configurations across every technology stack, ensuring accurate and up-to-date documents are created without draining technical resources. The AI engine thoroughly analyzes all components, their relationships, and patterns in legacy IT systems, Infrastructure Code, and Kubernetes clusters to provide a complete view of the IT landscape.

Intelligent Visualization & Documentation

AI-Powered Architecture Visualization

Content rich visualizations are generated accurately representing the deployed technology architecture. The intelligent engine creates interactive diagrams automatically mapping the relationships between components and services, which enables powerful drill-down capabilities. These visualizations offer exceptional clarity across diverse environments, allowing teams to comprehend complex architectures quickly and efficiently.

Comprehensive Documentation Creation

Beyond visualization, documentation on components, configurations, data flows, and relationships is comprehensive and detailed. The documents are created in required language (s) with accurate and consistent information which technical teams and business executives can consume in a format suitable for their use.

Version-Controlled Documentation

Integrating with existing CI/CD pipelines, documentation is version-controlled and evolves alongside every deployment change/ release. The code changes are automatically detected, triggering documentation update in line with the release for accuracy, while also maintaining the historical record of architectural evolution.

Verifiable and Queryable Architecture

The determined architecture and design are validated against enterprise-approved design patterns, security posture and adherence to organizational standards, and deviations are highlighted for prompt action in the release cycle. Additionally, the queryable nature of the documentation enables teams to extract specific insights or answer complex questions through a chatbot-like experience about the architecture, aiding teams with a better understanding both for faster development of new capabilities and managing the system.

Solution Architecture

Built as a flexible and extensible solution, it automatically generates visualizations and documents for all technology stack in an enterprise landscape, leveraging AWS technologies like Amazon Bedrock for AI, OpenSearch for RAG and S3 for data management, Step Functions for orchestration, ECS and Fargate for computation, CloudFront for delivery, Cognito for IDAM and Aurora for database needs, ensuring a secure, robust and scalable solution.

Fig 2

Processing Layer

Intelligent processing with AI and scaling with containerization

Technology Parser

Read the artifacts to generate understanding of the components and the relationship between them, from source code repositories.

Transformation Engine

Maps the identified components to the technology of the various providers, validates the identified architecture against approved design patterns, and generates intermediary output for final document generation by visualization engine.

Visualization Engine

Generate the diagrams and documents, index into the vector database and publish into repository. Also, as source artifacts change, automated processing ensures that vector indexing remains current.

Additionally, the RAG system plays a critical role across all processing components, for continuously learning from new documentation to improve future outputs.

Orchestration Layer

Manage all processing activities and maintain system state to ensure reliable & timely document generation with complete audit trail of activities performed.

User Experience Layer

Define the source systems, configure the integration with DevOps pipeline, inspect orchestration activities, and integrate with documents.

Source Control Integration

Source Control Integration - Connect to the defined source of truth of deployment enabling:

Solution Adoption – An example scenario

A global enterprise with a hybrid IT landscape spanning across on-premises servers, private cloud, and AWS & Azure. With both IaaS-based workloads and microservices-based applications, their current documentation is outdated. This resulted in delays and risks during the planning for any of their new transformation initiatives including a program to migrate the legacy customer relationship management (CRM) system to a modern cloud architecture.

Infosys SystemViz helped in documenting the existing IT landscape comprehensively including the CRM system, to create a pragmatic migration plan handling all integrations properly thereby increasing confidence in execution and reducing program risks.

Solution Walkthrough

Fig 3

Automated Pipeline mode

SystemViz integrates directly with CI/CD pipelines to maintain automatic synchronization between deployment configuration changes and documentation.

Fig 4

Interacting with the System

Source Configuration

The automated discovery and validation approach gives confidence that entire configuration code has been properly analyzed, be it modern cloud-native deployments, application settings, or hybrid environments containing legacy components, and calls out missing content/ configuration definitions. The validation feedback also flags file compatibility, and parsing errors if any.

Visualizations and Documentation Generation

With sophisticated graph algorithms and intelligent clustering techniques interactive architecture diagrams with contextual logical boundaries such as VPCs, resource groups, namespaces, application domains, or custom organizational patterns are generated.

The multi-perspective rendering system serves diverse stakeholder needs:

Richer content is brought into the architecture document with GenAI capabilities that include explanation of design decisions, component interdependencies, security implementations, application configurations, and compliance settings.

With interactive features like seamless navigation between visual diagrams and corresponding documentation sections, standards-compliant markdown document is generated. This enables export and integration with existing documentation platforms, wikis, and knowledge management systems that teams already use.

Query Interface and Enterprise Validation

The conversational AI with graph traversal algorithms enables semantic configuration queries, allowing users to ask intuitive questions like

"What components would be impacted by removing this database?" or "Identify all systems handling PII data?" or "Show me all configurations that reference this API endpoint."

The query engine with contextual understanding of component relationships and dependency chains provides specific component references and optimization suggestions that help technical teams understand impact scenarios and business stakeholders assess change implications.

Verifiable architecture starts with knowledge repository created with enterprise policies, standards & patterns and then continuous validation of derived architecture from configuration definitions. This validation system employs risk-based categorization algorithms powered by AI to identify deviations from defined standards and provides automated remediation suggestions and actionable corrective guidance.

The structured compliance reports are generated appropriately for different audiences—technical teams receive detailed remediation steps, while architecture review boards and compliance teams get executive-level summaries with clear risk assessments and corrective action priorities.

Conclusion

The Impact of SystemViz

Team Satisfaction

Delivery Performance

Operator Experience

Cost Efficiency

Security Posture & Risk Reduction

Auditing and Governance

Evolving the solution

With emerging GenAI tools like MCP Servers from AWS Labs on automated documentation generation for code repositories, the same can be adopted as the technology matures to enhance the capabilities of SystemViz.

We suggest integrating GenAI-assisted SystemViz solution into the software engineering process today to unlock its potential and navigate your technology landscape with confidence, delivering the comprehensive configuration intelligence that modern enterprises require.

References