Enterprise data modernization.
Without the enterprise overhead.
20 years building production data pipelines for retail, logistics, and enterprise operations. The same rigor and methodology — without the Big 4 overhead or the 18-month timeline.
The problems I get called about
200 SSIS packages that nobody wants to touch
Power BI reports that run for 10 minutes
Data scientists waiting 3 days for a data extract
No lineage, no documentation, no confidence in the numbers
A Fabric license you’re not using
Excel spreadsheets as the “source of truth”
I migrate in layers
I don't rip and replace. Each layer delivers value independently. You're never left mid-migration.
Bronze — Land everything
Raw, immutable, auditable. Every source system lands here with ingestion metadata, schema tracking, and append-only Delta tables.
Silver — Clean and conform
Deduplication, type enforcement, business-rule application, incremental processing. Schema contracts reject malformed data.
Gold — Analytics-ready
Dimensional models, semantic layer alignment, business logic centralization. Power BI Direct Lake for sub-second reporting.
The stack
What I've built
Loyalty points engine
Rewards accrual from raw POS across millions of customer interactions
Customer Master Profile
Identity resolution across CRM, ecommerce, POS, WMS into a single unified record
Ecommerce pipeline
Catalog, inventory, orders, allocations, receipts — JSON/XML to warehouse dimensions
GL reconciliation
POS/ecommerce/WMS financial events to general ledger posting and variance reporting
RFM segmentation
Behavioral scoring feeding marketing automation audience builds
Fabric modernization
Loop automation → Bronze, Materialized Lake Views → Silver, pipeline → Gold on Microsoft Fabric
Engagement types
Assessment
2 weeks
Current-state audit + technology roadmap
Medallion Build
8–16 weeks
Full Bronze/Silver/Gold implementation
AI Layer Add-on
4–6 weeks
NL query on your new lakehouse
Advisory Retainer
Ongoing
Fractional data engineering
Questions
Before you write.
Do I have to replace SSIS all at once?
No. I treat the estate as a set of obligations and move it in layers. A source can land in Bronze, be conformed in Silver, and feed a report from Gold while the package it replaces is still running.
What do you look at before a move to Microsoft Fabric?
Which packages land a source, which ones feed a report or a close, and which ones nothing reads. A Fabric license is not a sequence. The load-bearing feeds decide the order, especially where Power BI is what the business already trusts.
What does the first assessment cover?
About two weeks: the current state, the feeds that have to keep running, and a roadmap for where the medallion should sit. A full Bronze, Silver, and Gold build is a later scope, agreed after that picture exists.
Can the existing reports stay?
Yes, when people rely on them. The migration keeps those models fed. Direct Lake is useful once the grain and the refresh are understood. It is not a reason to start with the least-documented package.
How do I start?
Name the estate in ordinary language: where the packages live, which reports people use, and which source systems are in bounds. I do not need a connection string or an export to answer. I read the note and reply inside a business day.
Every engagement starts with a conversation.
No pressure, no pitch — just clarity on what's possible.
Start the conversation →