Architect Scenarios

Your SaaS product was just featured in a major newsletter.

You're the principal architect at a fintech startup.

You're setting up a new production environment for a web application with three tiers: a load balancer that accepts public traffic, application servers running your API, and a database.

You're deploying a three-service architecture: a web frontend (React SPA served from S3/CDN), an API server, and a PostgreSQL database.

You're building a real-time collaborative whiteboard app.

Your SaaS is growing globally.

Your 80-person engineering team is fully remote.

You're the founding engineer at a healthcare startup.

You've split your monolith into 3 microservices: an API, a worker, and a scheduler.

Your B2B SaaS stores customer data in an RDS PostgreSQL database and S3.

Your developer-facing public API processes 10 million requests per day.

Your consumer marketplace was hit with a volumetric DDoS attack last month — 50Gbps of UDP flood traffic that took the site down for 2 hours.

Your e-commerce platform is preparing for PCI-DSS Level 1 compliance.

Your engineering team of 60 uses a suite of internal tools: GitHub Enterprise, Jira, internal CI/CD dashboards, and a private npm registry.

You're building a blog platform that currently has 10,000 registered users.

You're building a consumer app where users upload profile photos and share image posts.

Your company generates 50TB of log and audit data annually that must be retained for 7 years for compliance.

You're building a document collaboration SaaS — think Google Docs for legal teams.

Your SaaS platform is expanding from US to EU.

Your B2B SaaS serves 200 enterprise tenants.

You're the architect at a financial services firm.

You just closed your Series A.

Your team has built a Node.

Your team is planning to deploy 5 microservices in containers.

Your team deploys to production 3–5 times per week.

Your platform has grown to 8 microservices communicating over HTTP.

After the Log4Shell incident, your security team has flagged container supply chain risk as a priority.

Your platform-as-a-service serves 30 enterprise tenants, each running containerized workloads on your shared Kubernetes infrastructure.

You're building a side project: a REST API that powers a mobile app you're making for fun.

Your data team runs nightly analytics jobs that join multiple large datasets in-memory.

You run an e-commerce site that holds daily flash sales.

Your data pipeline runs nightly batch jobs that process 200GB of raw event data into analytics-ready tables.

Your company runs a 10-year-old Windows Server 2012 application that handles internal HR workflows.

You're building the backend for a mobile game with 5M daily active users across North America, Europe, and Asia.

Your data science team has trained a sentiment analysis model using PyTorch.

You're building a lightweight backend for a contact form: receive a POST request, validate input, send an email via SendGrid, and log the event.

Users upload CSV files to your app for bulk data import.

Your Lambda-powered REST API serves a mobile app.

Your API handles 2 million requests/day today, growing 20% month-over-month.

You're building a data processing pipeline: (1) validate input, (2) enrich with third-party data, (3) run ML scoring, (4) write to data warehouse, (5) notify downstream systems.

Your platform processes user activity events — clicks, purchases, page views — at 1 million events per hour.

You're building an order fulfillment system.

You're building a new B2B SaaS product.

You're integrating with a customer's CRM via OAuth.

Three enterprise customers are ready to sign contracts — but each requires SSO.

Your platform has 12 microservices communicating with each other.

Your B2B SaaS currently stores all users in a single AWS Cognito User Pool with a tenant_id custom attribute.

Your engineering team of 20 all have broad AWS Admin access via IAM users.

You're redesigning the auth token strategy for a mobile app + web app + public API.

You run a B2B analytics SaaS with 40 enterprise customers.

You run a news website.

Your REST API serves a product catalog: 500,000 SKUs with prices, descriptions, and availability.

Your product team wants to run 5 concurrent A/B experiments: different checkout flows, homepage layouts, and pricing page copy.

Your company is building a developer platform with a public API.

Your e-commerce homepage must show personalized product recommendations based on a user's purchase history, geographic location, and active promotions.

You run a developer-facing public API with 3,000 registered API consumers.

Your data science team has trained a classification model in Python/scikit-learn.

Your ML team has built a churn prediction model.

You're building a RAG (Retrieval-Augmented Generation) app that lets users query a corpus of 2 million internal documents.

Your company wants to build an AI assistant for customer support that knows your product inside-out.

Your ML team wants to fine-tune a 7B parameter open-source LLM (Llama-3-8B) on 50GB of proprietary text data.

You're building a real-time product recommendation engine.

Your product has an AI writing assistant powered by an LLM.

Your web app sends welcome emails, processes image thumbnails, and syncs data to a third-party CRM after user signup — all currently happening synchronously in the API response.

You're processing payment events from a webhook.

You're designing the messaging layer for an e-commerce order processing system.

Your payment processing pipeline uses SQS → Lambda.

When a user places an order, 20 downstream systems need to react: inventory deduction, shipping calculation, loyalty points, fraud scoring, analytics, 5 regional notification services, 3 ERP integrations, and 7 partner APIs.

You're the architect for a financial trading platform.

After an acquisition, your company now has two engineering teams: one building on AWS (SQS, Lambda), one building on Azure (Service Bus, Azure Functions).