1.3 Mn
Monthly users/sessions
76K+
Monthly transactions
40+
Countries served
Industry
Sub-Industry
When Inferenz first engaged the clien t, the entire data operation ran on on-premise SQL Server with queries taking hours, reports built in Excel and Sisense, and marketplace data flowing through a costly third-party subscription. Over the last few years , Inferenz delivered a phased modernization program, migrating to Snowflake, replacing vendor dependencies with in-house Python APIs, enabling self-serve Power BI reporting, building Marketlo for in-house customer segmentation, and positioning the platform for AI and predictive analytics from 2025.
During project inception, the client's infrastructure had outgrown its foundations, on-premise, vendor-dependent, and unable to support the reporting speed or analytical flexibility a scaled e-commerce operation needed.
SQL Server held hundreds of terabytes of e-commerce data. Queries powering daily business decisions took hours. The system could not scale further without capital investment that still would not match cloud performance.
Teams across merchandising, marketing, and analytics worked from Excel and Sisense dashboards that took 3-4 hours to refresh. Every change required a data team request — decision-making was slow and team-dependent.
A paid Data Virtuality subscription pulled marketplace and ad platform data from Amazon, Adobe, and Google. It was expensive, unreliable under load, and left the team entirely dependent on vendor infrastructure for business-critical data.
Customer segmentation and ranking ran through a third-party vendor, taking 3-4 days per cycle. The team had no control over logic or turnaround, creating recurring delays in merchandising and marketing decisions.
Inferenz modernized the entire data stack from on-premise SQL Server to a cloud-native Snowflake platform, eliminating vendor dependencies, enabling self-serve reporting, and building the foundation for AI and predictive analytics.
The entire data operation migrated from SQL Server to Snowflake, with Pentaho ETL jobs moved in parallel and no disruption to existing flows. Data Virtuality was replaced with Python APIs pulling Amazon, Adobe, and Google data directly into Snowflake, eliminating the subscription cost and bringing the pipeline fully in-house.
Sisense and Excel were replaced with Power BI following a competitive analysis of POC . SSAS cube logic was re-engineered into Snowflake and Power BI, cutting analytical processing from 3 hours to 45 minutes. Business teams were enabled to build their own dashboards independently from day one.
Inferenz then built Marketlo, an in-house customer segmentation and ranking solution on Snowflake, replacing the third-party vendor entirely. Turnaround dropped from 3-4 days to approximately 1 day, with full ownership of segmentation logic now sitting with the client.
The project expanded into advanced analytics, API integrations, and a Natural Language Analytics agent POC. The Snowflake platform built now serves as the foundation for AI and predictive intelligence development.
Cloud migration achieved
The entire data operation: pipelines, ETL jobs, integrations, and analytics, migrated to Snowflake completely.
Integration cost reduction
Data Virtuality subscription replaced by Python APIs — recurring vendor cost eliminated; pipeline stability improved.
Customer segmentation turnaround
Segmentation cycles reduced from 3-4 days to ~1 day via Marketlo, built in-house on Snowflake.
Business reporting
Merchandising, marketing, and analytics teams build and manage their own Power BI dashboards independently.
Whether you’re starting with data modernization or exploring AI copilots, we’re here to help.
Contact Us
