Alpura, a leading dairy company in Mexico with over 10,000 employees including 5,500+ field sales representatives, is transforming from a traditional dairy producer into a data-driven nutrition enterprise. Facing fragmented data sources, inconsistent metrics and slow decision cycles across their commercial, supply chain and factory operations, Alpura built manchasONE — their unified data intelligence platform powered by Databricks. In just 120 days, they established a governed lakehouse foundation and began deploying apps that deliver real-time insights across the organization.
Fragmented data blocking speed and scale across dairy operations
For decades, Alpura operated like most traditional food and beverage companies — relying on disparate systems, spreadsheets and manual processes to run their business. With over 10,000 employees spanning dairy production, revenue growth, marketing decisions, distribution and retail partnerships, the company faced mounting pressure to make faster, data-driven decisions. Yet their technology landscape had become a barrier rather than an enabler.
“We had Snowflake, Matillion, AS/400 extracts, Oracle ERP migration to Oracle Fusion Cloud, spreadsheets, Power BI and scattered business rules,” Marvin Nahmias, Chief Transformation, Data and AI Officer at Alpura, explained. “No catalog, no governance, no semantic layer — siloed decisions.” Business teams across the entire organization all needed access to reliable data, but couldn’t get it in time to make informed decisions. Excel files proliferated as workarounds, creating version control nightmares and conflicting metrics.
The technical challenges ran deep. Alpura’s fragmented tooling created duplicated logic, high infrastructure costs and little semantic consistency. Business decisions lacked a reliable single source of truth. Without unified definitions for key performance indicators, different teams would arrive at conflicting conclusions from the same business questions.
The consequences were clear: Deployment cycles stretched 12–16 weeks from prototype to production, business intelligence (BI) initiatives couldn’t scale and critical use cases remained unbuilt. Alpura couldn’t deploy real-time OEE (overall equipment effectiveness) dashboards for factory optimization, AI copilots for commercial teams, IoT (Internet of Things) applications for quality control or predictive maintenance models. The company’s ambition to double their enterprise value within four years necessitated a fundamental transformation of their data architecture.
Building manchasONE: A unified platform centered around apps
Alpura made a decisive move to consolidate their entire data, business rules (semantic layer), AI, BI and application infrastructure onto the Databricks Data Intelligence Platform. They called their new unified platform “manchasONE” — named after their mascot, Manchas the cow, with “ONE” representing everything unified in a single trusted platform. The transformation focused on three priorities: establishing data governance through Unity Catalog, centralizing their semantic layer with consistent business rules and metrics, and enabling rapid application development through Databricks Apps.
The foundation came together remarkably fast. “We completed the foundational transformation — lakehouse, medallion, governance and some semantic layers — in the first 120 days,” Marvin shared. The team implemented a medallion architecture bringing data from their Oracle ERP, AS/400 systems and external IoT sources from factory sensors into Delta tables under Unity Catalog governance. SQL warehouse provided the semantic layer where business rules, definitions and metrics all lived together, accessible to all teams.
Databricks Apps became the front door to manchasONE. The first application addressed a critical operational need: an IoT Sensors app for a major dairy lab that monitors temperature and quality parameters in real time. The team built custom Arduino sensors that connected directly to Databricks streaming tables, capturing data every five seconds. This approach provided significant cost savings while giving the company full control and flexibility over their monitoring infrastructure.
The second app delivered immediate commercial value: a near real-time decisioning app fully integrated with Walmart Scintilla. Alpura’s retail partnerships generate massive transaction volumes, but decisions at the time often relied on stale data refreshed only every few hours.
The new app enabled teams to make data-driven decisions based on current inventory positions and sales trends, while running predictive what-if scenarios for strategic retailer conversations.
Apps deliver insights to hundreds, from the factory floor to field sales
Beyond cost savings, the most significant improvement came from centralizing and governing Alpura’s semantic layer inside manchasONE. “The most transformative thing is speed, governance and intelligence unified,” Marvin emphasized. “Apps in days, not months.” Business teams now work with consistent definitions, metrics and business rules — no more conflicting KPIs or spreadsheet chaos. Alpura launched a formal governance office to maintain data quality and semantic clarity as the platform scales.
The transformation has fundamentally changed how Alpura operates and makes market decisions. Deployment cycles were reduced from 12 weeks to just two to five days, enabling the team to build over 10 apps that serve hundreds of users across the organization. Real-time commercial execution has become possible. Field sales representatives can now access current inventory positions, customer insights and pricing guidance through governed apps rather than outdated reports. Factory operations teams monitor OEE dashboards and receive automated alerts when production metrics deviate from targets, with predictive maintenance as a key use case. Supply chain teams leverage AI-enhanced demand forecasting models deployed through Model Serving, enabling more accurate planning and reduced waste.
“Our vision is to deploy AI apps, agents and Genie spaces for power users, accelerating tactical work, enabling predictive supply chain capabilities, powering IoT factory intelligence and providing a commercial cockpit integrated with RGM [revenue growth management] tools to prioritize activities for our 5,500-person sales force,” Marvin outlined. With manchasONE as their unified data intelligence foundation, Alpura is positioned to innovate rapidly and compete as a modern, data-driven nutrition company.