Case Study Spotlight:

Global Consumer Commerce Organization Modernizes Product and Pricing Platform on AWS
Industry: Consumer Products, Furniture, Omnichannel Commerce
Region: Global
Solutions: Application Modernization, Data and Analytics, Managed Services
AWS Partner: An AWS Consulting Partner specializing in migration and modernization

About the Customer:

The customer is a global furniture and home furnishings company with a broad product portfolio across several brands and regions.

Product and pricing information originates from multiple systems with different formats, attributes, and update cycles. These differences made it difficult to keep downstream channels aligned on current catalog content and price across markets and sales channels.

01. The Challenge:

During joint discovery, the customer and Initech identified several recurring issues in the existing product and price publication process:

  • Product and price data was stored in several source systems, each with its own structure and extract process.

  • Updates typically flowed through nightly file based feeds and manual adjustments, which introduced lag and inconsistency.

  • Search and browsing experiences for product data were limited by the underlying storage and indexing tools.

  • The existing publication tooling was not designed for multi region use, complicating global availability and recovery plans.

  • Maintaining custom scripts, one off integrations, and on premises infrastructure consumed engineering time that could have been used for higher value work.

To address these issues, the teams agreed to design a centralized product and pricing platform on AWS that could:

  • Act as a consistent source of truth for product and pricing data.

  • Support a range of current and future consumers across digital, retail, and partner channels.

02. The Solution:

Initech and the customer co-designed the platform, guided by an internal technical architecture for the initiative. The solution is structured around three main stages.

1. Extract and Staging

  • Product and pricing data from systems such as CSP and XML based feeds is loaded into an Amazon Aurora PostgreSQL staging database on AWS.

  • AWS Lambda, AWS Step Functions, and container based extract tasks handle full, incremental, and on demand loads.

  • Data is partially denormalized at this stage and tagged with extract identifiers for traceability.

2. Transform and Load

  • From staging, serverless pipelines orchestrated by Step Functions apply validation, enrichment, and mapping into a domain model optimized for read performance.

  • Product and price are treated as separate domains and versioned independently.

  • The transformed data is written to Amazon DynamoDB tables.

  • Changes in DynamoDB trigger updates to Amazon OpenSearch Service indexes that support search and catalog access patterns, including furniture specific attributes and configurations.

3. API and Consumption

  • Amazon API Gateway exposes APIs that provide consistent access to products, options, prices, and validation results.

  • Back end implementations use Lambda and container based services written in TypeScript for performance sensitive paths.

  • Authentication and authorization integrate with existing identity providers.

High Availability and Multi Region Design:

  • DynamoDB global tables and OpenSearch replication keep data available across regions.

  • Traffic routing and failover are tested using planned failure scenarios.

  • Monitoring and diagnostics use Amazon CloudWatch along with OpenSearch based audit indices to track ingest, transformation, and API usage.

Ways of Working:

Throughout the project, Initech and customer engineers worked from a shared backlog:

  • Initech focused on AWS architecture, infrastructure as code, and pipeline design.

  • Customer teams contributed domain knowledge, validation rules, and integration with upstream and downstream systems, including furniture catalog tools and pricing engines.

03. The Outcomes:

The platform has changed how product and pricing data flows through the customer environment:

  • Product and price updates move from sources to consuming systems with less manual intervention and more predictable timing.

  • Downstream channels now consume a single, well defined representation of product and pricing data, which reduces discrepancies across tools, regions, and sales channels.

  • Search and catalog experiences draw on OpenSearch indexes designed for common access patterns, improving usability for internal merchandising teams and external shoppers.

  • Operational tasks such as monitoring ingest health, investigating failures, or reprocessing data are now handled through AWS based dashboards, alerts, and runbooks instead of ad hoc scripts.