All projects

Live · 2024

Optisage

Backend for an AI-driven analytics platform for Amazon sellers — Selling Partner API integrations, processing, and Redis-backed workflows.

Backend engineer

  • Laravel
  • MySQL
  • Redis
  • Docker
  • Amazon SP-API
Live site

Problem

Seller analytics needed reliable ingestion from Amazon SP-API, then processing and caching so dashboards were not waiting on live marketplace calls.

What I built

  • Implemented SP-API integrations and data-processing workflows in Laravel.
  • Used Redis for application caching and job-backed analytics paths.
  • Containerized the service with Docker for repeatable deploys.

Architecture

flowchart LR
  amazon[Amazon SP-API] --> ingest[Laravel ingest]
  ingest --> jobs[Redis jobs]
  jobs --> store[MySQL]
  store --> api[Analytics API]
  api --> app[Seller dashboards]

Outcomes

  • Working seller-analytics backend that pulls and processes Amazon marketplace data.
  • Redis-backed paths for application functionality that would otherwise hit SP-API on every request.

Context

Optisage is an analytics platform for Amazon sellers. I built backend functionality around Amazon Selling Partner API integrations, data processing, and Redis-backed application paths.

Seller dashboards cannot wait on live SP-API for every chart. The useful backend is the one that ingests, stores, and serves derived data on a clock you control.

Ingest and process

Laravel owned the HTTP surface and the workers. Incoming SP-API data was processed into analytics-friendly records in MySQL. Redis sat in front of hot application reads and as a queue for jobs that would otherwise block a request.

Docker kept the service reproducible. The interesting design work was the boundary: what is fetched from Amazon, what is stored, and what is safe to cache.

This is a backend case study, not a claim about the AI product on top. The job was reliable integrations and workflows sellers could actually load.