Automated OTIF monitoring and conversational AI analytics for a molten metal supply company

Company overview

The client operates in the aluminum recycling and liquid metal supply industry, serving high-pressure die-casting foundries across multiple locations. The industry is highly logistics-driven, requiring precise coordination and timely deliveries, as molten metal has limited holding time. Any delay can disrupt production, making reliability, efficiency, and supply chain visibility critical for operations.

Process digitization
Workflow automation
Legacy modernization
Industry
Molten metal supply
Total employees
500
Engagement area
Applications Engineering

Core business challenges

  • Manual way to analyze which machine was running low
  • No-Metal event root cause hard to establish across systems
  • Operational data siloed across sensors, weighbridge, and GPS
  • No analytical layer converting raw data into insight
  • No data analyst to consolidate operational reporting

Technical challenges

  • Three disconnected source systems (sensors, weighbridge, GPS) with no shared data model or event correlation
  • Minute-level sensor data arrived with duplicates, zero-level anomalies, and spikes requiring noise filtering
  • No-Metal detection needs three simultaneous conditions, and supply-vs-internal attribution needs minute-accurate GPS overlap
  • Vehicle-unavailability windows across 1,440 daily minutes and OTIF % both require dynamic denominators to stay accurate

Teams impacted
Operations
Site supervisors
Quality assurance
Management

Solution

OPTIMA operational data platform

A SQL Server dimensional model processes all three source systems through a 14-step ETL pipeline into 11 fact tables, with full-reload and incremental daily processing.

OTIF Bot — conversational AI analytics

A secure, on-premises multi-agent assistant lets authorised users query the database in natural language and receive text, tables, and charts, retaining context across a conversation.

Built around existing data sources

No operational system was replaced; OPTIMA extends at the process and experience layers, preserving data ownership and protecting operational continuity.

Outcomes

6
Automated Cause Classification
3 to 1
source systems unified into one operational data model
11
fact tables built via a 14-step automated ETL pipeline
2
Early Detection of Sensor Issues
24×7
anomaly detection at scale via natural-language query
100%
Complete Traceability & Audit Trails
6
Automated Cause Classification
3 to 1
source systems unified into one operational data model
11
fact tables built via a 14-step automated ETL pipeline
2
Early Detection of Sensor Issues
24×7
anomaly detection at scale via natural-language query
100%
Complete Traceability & Audit Trails

Bottom - line impact

Advaiya made service-failure time visible by machine, cause, shift, and customer, automated OTIF collation, and removed the data-analyst dependency for routine insight for CMR OTIF customers.

Featured technologies

Testimonial

Advaiya’s solution significantly improved the visibility for OTIF (On-Time In-Full) performance, helping us to identify no-metal time and supply chain issues for our customers

IT – Marketing Executive

Recent case studies

Ready to revolutionize your business?