☁️ Solution Analyzer for Azure

← Back to Main Page

Help & Documentation

Everything you need to know about Architect's Time Saver

What is Architect's Time Saver?

Architect's Time Saver (ATS) is an AI-powered tool designed to help Azure Solutions Architects rapidly evaluate requirements and produce detailed architecture designs for cloud solutions on Microsoft Azure.

Given a natural-language description of your requirements, ATS analyses the demand, selects appropriate Azure services and SKUs, estimates monthly costs, generates an architecture diagram, and provides concrete improvement suggestions — all in one automated workflow.

ATS does not replace the architect. It removes the repetitive, time-consuming groundwork so you can focus on higher-level decisions, client conversations, and final review.

Who is it for?

ATS is built for professionals who design and evaluate Azure solutions on a regular basis:

  • Azure Solutions Architects — quickly produce a first-draft architecture and cost estimate for a new customer requirement or proof-of-concept.
  • Pre-sales & Consulting Engineers — generate a credible architecture skeleton during or right after a discovery call.
  • Cloud Leads & Technical Managers — validate feasibility, compare design options, and understand rough cost implications without deep hands-on analysis.
  • Architects learning Azure — explore how Azure services fit together and understand trade-offs across different design patterns.

Quick Analysis vs Full Analysis

ATS offers two modes, so you can choose the right balance between speed and depth.

Quick

Quick Analysis

Produces an initial high-level assessment of your requirements: assumed Azure resources, architecture type, design decisions, trade-offs, and a confidence level.

  • ✅ Runs in seconds to a couple of minutes
  • ✅ Great for early exploration or rapid scoping
  • ✅ Identifies the right Azure service families
  • ✅ Surfaces key design decisions and trade-offs
  • ⬜ No exact SKU selection or pricing
  • ⬜ No architecture diagram
  • ⬜ No security threat model

Best when: you need a fast sanity-check or want to validate whether a requirement is worth a full analysis.

Full

Full Analysis

Runs the complete six-step agentic workflow: requirements analysis, service & SKU selection, security analysis, design validation, diagram generation, cost calculation, and improvement tips.

  • ✅ Exact Azure service and SKU recommendations
  • ✅ Live pricing from the Azure Retail Prices API
  • ✅ Auto-generated architecture diagram
  • ✅ STRIDE threat model + security baseline
  • ✅ LLM-judge quality validation
  • ✅ Actionable improvement suggestions
  • ⏱️ Typically takes 5–10 minutes

Best when: you need a complete, reviewable design artefact — for a proposal, workshop, or architecture review.

How to get the best results

Write a clear, detailed requirements description

The richer your description, the more accurate the output. Include the business context, expected scale (users, transactions, data volume), availability requirements, compliance constraints, and any existing services you must integrate with.

Example — less effective:

"I need a web app with a database."

Example — more effective:

"A customer-facing e-commerce portal for a retail company, expected 10 000 concurrent users at peak, 99.9% uptime SLA, GDPR-compliant data storage in West Europe, integration with an existing on-premises ERP via VPN, and a managed relational database with read replicas for reporting."

Attach relevant documents or diagrams

If your model supports vision or document processing, you can upload existing architecture diagrams, specification documents, or screenshots. ATS will incorporate them into its analysis and produce a summary of what it found in the uploaded files.

Choose Quick or Full Analysis

Start with a Quick Analysis if the requirements are unclear or you just need a first opinion. Run a Full Analysis when you want a complete, export-ready artefact.

Pick the right model

Use the Model selector to switch between AI models. Faster models are great for Quick Analysis while smarter models produce better results for Full Analysis. See Choosing a model below for details.

Review and iterate

ATS output is a starting point, not a final answer. Review the design decisions and trade-offs, check the cost breakdown, and run the analysis again with a refined description if needed. Use the improvement suggestions to guide follow-up iterations.

Export your results

Once satisfied, export the full analysis to PDF, Word, or Markdown for further refinement.

Choosing a model

ATS supports multiple AI model providers and models. You can switch the active model using the Model selector below the requirements text area. The choice of model affects both the speed and the depth of the analysis.

Fast

Faster models

Lightweight models optimised for speed and cost efficiency. They produce solid results in less time and are a good match for Quick Analysis or rapid iteration.

  • GPT-5.x mini
  • Mistral Small

Best when: you need a fast turnaround, are exploring ideas, or running multiple Quick Analyses in succession.

Smart

Smarter models

Larger, more capable models that produce richer design decisions, better security analysis, and more nuanced trade-offs. They take longer but are a better fit for Full Analysis.

  • Claude Sonnet
  • GPT-5.4
  • Gemini Pro

Best when: you need a complete, high-quality architecture artefact for a proposal, workshop, or architecture review.

You can mix and match: run a Quick Analysis with a fast model to validate direction, then switch to a smarter model for the Full Analysis.

Tips & best practices

  • Be specific about region and compliance. Mentioning a target Azure region (e.g. West Europe) and any compliance requirements (e.g. GDPR, PCI-DSS, ISO 27001) helps ATS select region-available services and apply the appropriate security baseline.
  • Include non-functional requirements. Availability SLA, RPO/RTO targets, expected latency, and throughput numbers directly influence service and SKU recommendations.
  • Mention your team's constraints. If your team has limited Kubernetes expertise, or prefers PaaS over IaaS, include that — ATS will factor it into design decisions.
  • Always review AI output with a qualified architect. ATS is a productivity tool, not a replacement for expert judgment. Pricing estimates are indicative, service availability can change, and AI models can occasionally make mistakes. Treat the output as a high-quality first draft.
  • Use the security analysis section. The Full Analysis includes a STRIDE threat model and Azure Security Benchmark alignment assessment. These are valuable starting points for security conversations with your customer.
  • Check the improvement suggestions. The improvements section at the end of a Full Analysis often surfaces cost-saving opportunities (e.g. Reserved Instances, right-sizing) and reliability patterns you may not have considered.