Project
Aegis — Amazon/E-commerce Automation SaaS
Aegis is a multi-tenant SaaS platform that automates a full slice of Amazon/e-commerce operations around high-demand products, limited releases, collectibles and Pokémon cards: managing an account fleet, requesting invitations for targeted products, detecting important emails, tracking orders, triggering configured auto-buy flows and centralizing activity in a client/admin dashboard.
The product is live in production and used by several clients. I designed the architecture and built the dashboard, API, workers, persistent job queue, Stripe/Discord integrations, messaging layer, monitoring and deployment infrastructure. The result is a real product used day to day, with concrete reliability, security, quota, UX, support and production maintenance constraints.
Key facts
Real SaaS product running in production.
Used by several clients through a monthly subscription model.
E-commerce use case: high-demand products, limited releases, collectibles and Pokémon cards.
Client/admin dashboard, messaging, event tracking and monitoring.
Backend automations with NestJS API, separate workers and persistent jobs.
Stack: TypeScript, Next.js, NestJS, PostgreSQL, Prisma, Docker, Nginx, PM2.
GitHub Actions CI/CD, healthchecks, backups, rollback and self-hosted infrastructure.
Frequently asked questions
What is Aegis used for?
Aegis centralizes and automates Amazon/e-commerce workflows for targeted products, events, emails, orders and backend jobs inside a client/admin dashboard.
What is the e-commerce use case?
The product covers operations around high-demand products, limited releases, collectibles and Pokémon cards, with monitoring, statuses and configured automations.
What technical stack does it use?
The main stack includes TypeScript, Next.js, React, NestJS, PostgreSQL, Prisma, Docker, Nginx, PM2, GitHub Actions and CI/CD.
What does this project demonstrate?
Aegis demonstrates real product experience: production SaaS, subscribed clients, multi-tenant architecture, workers, monitoring, self-hosted deployment and operational maintenance.
Problem
Clients had to manually track scattered Amazon operations: accounts, targeted products, invitations, emails, orders, job statuses and incidents. As volume grows, manual tracking becomes slow, fragile and impossible to maintain cleanly for several clients.
Tech Stack
Key Features
Aegis is not a mockup: it is a production platform used to manage Amazon/e-commerce workflows end to end, including collectibles and Pokémon cards.
Client/admin dashboard
Next.js interface with client space, admin view, subscription management, settings, product collection, messaging and live operations tracking.
Account fleet and backend jobs
NestJS API and separate worker to manage long-running operations: persistent job queue, error recovery, result history, quotas and API/worker separation.
Product invitations and Amazon tracking
Invitation requests for targeted products, email response detection, per-account statuses, order tracking and client-side activity consolidation.
Auto-buy and orders
Configured auto-buy triggering, order tracking, payment/delivery statuses and incident reporting in the interface.
Monitoring and operations
GitHub Actions CI/CD, blue/green deployment, healthchecks, production doctor, service status, encrypted backups, restore drill and rollback.
Self-hosted network infrastructure
Design of a private network layer with VPS, VPN tunnel, self-hosted proxies, outbound flow supervision, reverse proxy, firewall and hardened public exposure.
Solutions & Results
I built a SaaS console that turns these operations into job-driven workflows: each client has an isolated space, while the admin manages the fleet, requests, messaging, catalog, orders and monitoring. The infrastructure is self-hosted with VPS, Docker, Nginx, PM2, CI/CD, backups, rollback, healthchecks, operational supervision and a private network layer with self-hosted VPN/proxies to control outbound flows.