Sripathi Mohanasundaram
Case study — Fractal Analytics

Synapse → Microsoft Fabric OneLake migration

Migrated data from Azure Synapse to Microsoft Fabric OneLake for a global retail client, ensuring integrity and optimizing storage architecture — while monitoring and optimizing pipelines to achieve 99.9% uptime through the cutover.

Role
Data Engineer — pipeline design, migration
Period
Jan 2023 — Present
Stack
Azure Synapse · Microsoft Fabric OneLake · Databricks · Azure Data Factory · SQL
Outcome
99.9% uptime through cutover

The problem

The retail client's warehouse ran on Azure Synapse with metadata pipelines that needed to scale further, and 50+ transactional tables were still coupling downstream reporting directly to the operational system — adding load and slowing everything behind it.

Approach

I contributed to the ODS project, replicating those 50+ transactional tables to take load off the source system and enable faster scaling. From there the warehouse itself moved onto Microsoft Fabric OneLake, with the migration built to preserve data integrity and optimize the storage architecture rather than lift-and-shift the existing structure as-is.

Architecture diagram

Azure Synapse → Microsoft Fabric OneLake, with Databricks and Azure Data Factory driving ingestion throughout.

Results

Uptime
99.9% maintained across pipeline operations, supporting rapid scaling for high-volume data.
Scalability
Replicating 50+ transactional tables reduced system load and enabled faster scaling for downstream consumers.
Integrity
Migration preserved data integrity while optimizing the underlying storage architecture.
Stakeholders
Worked directly with stakeholders to design scalable solutions delivering measurable outcomes aligned with growth goals.

All work

Back to the index of collections and case studies.