Amazon Service Provider

Automation for a full-service Amazon agency. Variation checks and case creation run by bots, and an account-level view that shows a brand exactly how its sales are performing.

Services
RPA development
Process automation
Industry
B2B · eCommerce and marketplace services
Case study cover: an illustration of robotic process automation, with a robot and AI chips connected around the letters RPA

Overview

Together, Sphinx and an Amazon Service Provider

Our client is a Chicago agency that has been running end-to-end services for brands on Amazon since 2015: supply chain, product development, eCommerce, marketing and market strategy, for companies who want to grow on the marketplace rather than merely appear on it.

They came to Sphinx Solutions for process automation on Amazon, because the thing holding the service back was a lack of transparency.

What they offer spans five areas. Distribution: logistics, warehousing, inventory and imports. Sales: performance reporting, cost-of-goods control and a partnership network. Marketing: market gap analysis, web optimisation, photography and content. Product development: design insight, manufacturing compliance and cost reduction. Sourcing: access to more than 150 manufacturing partners, quality checks and packaging to global standards.

Running all of that across many brands means a great deal of repetitive checking, and the parts that were manual were the parts nobody could see into.

They needed variation checks and case creation to happen easily and consistently, which is precisely the shape of problem RPA is for.

Variation Checks and Case Creation, Run by Bots

The challenge

The client wanted to give brands long-term success on the largest marketplaces, and the obstacle was transparency: too much of the work happened in places the client and the brand could not see into.

Variation checks and case creation had to become easy, repeatable and visible rather than manual and occasional.

Before building anything we went through the requirements and the business goals with them, because in marketplace operations the process is the product and getting it wrong automates the wrong thing faster.

From there we worked the concept into an ideation process and a strategy, then chose the tools to match rather than the other way round.

The solution

We delivered analytics and automation together, so the brand can see the position and the agency can act on it.

  1. An executive account-level view, giving an exact picture of how a brand’s sales are performing rather than a report assembled after the fact.
  2. More than fifteen automated processes on Amazon, covering the repetitive checking that used to be done by hand.
  3. Built on UiPath, using it as the whole automation platform: embedded analytics, process mining, SaaS-based RPA and test automation in one place.
  4. High accuracy first, because in marketplace case creation a wrong case costs more than a slow one.
  5. Delivered through the three engagement models we run, RPA consulting, end-to-end development and maintenance, so the automation is supported rather than handed over.

The transparency was the point. Automating the checks is what let the agency show a brand what was happening to its account.

Customer testimonial

“We feel highly privileged to have Sphinx Solutions as our technology partner. They have an excellent team of industry professionals who helped us at every stage of the project. We are satisfied with what they delivered. Highly recommended.”

Tech stack

What It Was Built With

Questions &
Answers

What people usually ask about Amazon Service Provider and about building something like it.

A business that runs the parts of selling on Amazon a brand does not want to run itself: distribution and logistics, sales reporting, marketing and content, product development and sourcing. The good ones make a brand’s position on the marketplace visible as well as profitable.

It depends on the platform. UiPath builds its workflows in VB.NET and C#, Blue Prism is C#, and Automation Anywhere uses its own scripting alongside Java and Python. Python turns up almost everywhere for the glue work. Much of RPA is configured rather than coded, which is the point of it, but the extension points are where a real language is needed.

Rule-based, repetitive work with a clear right answer: invoice data extraction, accounts payable, data entry and reporting, payment processing, and reconciliation of any kind. The test is whether a person following the rules would always reach the same result.

Connect with us

Turn Your Ideas Into Reality. Let’s Connect and Create Something Groundbreaking.

  • USA
  • UK
  • UAE
  • India

Enter your contact details

We’ll get back to you in 1 to 2 business days.