Who Made This

The AI Solutions Wiki is built and maintained by Linda Mohamed , an AWS Community Hero and AI Solutions Architect based in Austria. Linda works with teams across Europe building production AI systems on AWS - from first prototype to governed, observable, production-grade deployments.

If you want to work together:

What This Wiki Is

The AI Solutions Wiki is a practical knowledge base for teams building AI-powered products and workflows. The focus is on applied AI - not theoretical research, not vendor marketing, but the real patterns, tools, and decisions that come up when you are turning an idea into a working system.

Every article here is written to be immediately useful. Definitions include context about when and why you would use something, not just what it is. Guides include real trade-offs. Comparisons are honest about where each option falls short.

Who It Is For

This wiki is useful at two career stages:

Practitioners who are new to AI - engineers, product managers, and architects who understand software but are mapping a new terrain. The glossary, guides, and frameworks sections give you the vocabulary and structure to navigate AI decisions confidently.

Experienced AI builders - people who have shipped AI systems and want a reference for patterns, tools, and comparisons without re-reading vendor documentation. The patterns, case-patterns, and comparisons sections are aimed at you.

The content is also useful for anyone who needs to explain AI decisions to stakeholders - the frameworks and comparisons are written to support clear communication, not just technical implementation.

How the Content Is Organized

The wiki is organized into three areas: learn, build, and reference.

Learn

Basics : 15 “What is X?” articles for people starting from zero. Plain-English explanations of computers, databases, APIs, Git, and the terminal before introducing AI.

Levels : A five-level structured curriculum from hardware fundamentals (Level 0) to building production AI systems (Level 4). Each level is two to four articles. Designed to be read in sequence, but every article works as a standalone reference.

Foundations : The engineering principles that apply to every AI system regardless of which tools or models you use. Covers SOLID, Clean Architecture, Domain-Driven Design, testing strategy, CI/CD, and the AWS Well-Architected Framework.

Learn Through Your World : Persona-based learning paths that explain software and AI concepts through metaphors from your existing domain. Currently active: The Juggler (distributed systems), The Fashionista (version control and deployment), The Craftsperson (system design and production quality).

For You : Audience-specific reading paths for product managers, founders, consultants, finance professionals, vibe coders, and students. Each path cuts across the wiki to surface what is most relevant for your role.

Build

Guides : Step-by-step implementations. 174 guides covering RAG systems, async job queues, multi-agent pipelines, sprint planning with AI, and production deployment patterns.

Patterns : Reusable architectural patterns that appear across AI projects: RAG, agent orchestration, data pipelines, prompt design, evaluation harnesses.

Case Patterns : Deployment patterns drawn from real production use cases. Anonymized, with details focused on what makes the architecture work and where it fails.

Architecture : Visual reference for the architectural patterns that underpin AI systems.

Reference

Tools : 175 tools covered. What each tool does well, what it does not, pricing traps, and when to choose it over alternatives.

Comparisons : Side-by-side analysis for 74 common decision points: Bedrock vs Azure OpenAI, RAG vs fine-tuning, Claude vs GPT-4, Airflow vs Step Functions.

Frameworks : Structured thinking tools for AI governance and planning. EU AI Act, ISO 42001, OECD Principles, Team Topologies, Wardley Mapping.

Glossary : Plain-English definitions for 442 AI, ML, and cloud terms. Every definition includes context on when and why you would encounter that concept.

Solutions : Industry-specific AI applications covering finance, insurance, media, logistics, healthcare, and more. Each article describes the problem, the architecture, the tools, and the operational considerations.

Ideas : 80 short concept pieces for automation use cases worth exploring before committing to a build.

AI Tools and Editorial Process

This wiki is curated with the assistance of AI tools. Content is researched, drafted, and reviewed using Claude (Anthropic), Kiro (AWS), and ChatGPT (OpenAI). Images are generated with NotebookLM (Google). Animations and visual assets are produced with Canva.

All content is reviewed by Linda Mohamed before publication. The use of AI tools does not reduce editorial responsibility. Every article reflects decisions made by a human editor. Errors, omissions, and updates are the responsibility of the site maintainer.

Adjacent Resources

The wiki is part of a small ecosystem of free and paid resources for technical practitioners and freelance engineers shipping AI:

  • Freelancer Templates : production-grade contracts, proposals, statements of work, briefs, and project documentation templates for AI and software freelancers. The reference set used in client engagements.
  • Freelancer Automation : automation recipes, AI playbooks, and workflow examples for solo operators and small consultancies. Practical patterns for billing, intake, scoping, and delivery automation.
  • AI Workshops Online : structured workshop programmes that take teams from first prototype to production AI on AWS.
  • Always Ahead Newsletter : weekly notes on AI engineering, governance, and the state of practice.

These are run by the same author and link back to the wiki for canonical technical reference.

How the Content Is Maintained

Articles are reviewed and updated as tools evolve and new patterns emerge. Pricing information and model capabilities change frequently, always verify current pricing at the provider’s official documentation before making architecture decisions based on figures quoted here.

If you find an error or a gap, the best way to contribute is through the GitHub repository at github.com/linda-mhmd/ai-solutions-wiki .