LuxembourgChemicals and materials2025 – ongoingNexa
Predictive process models and a contract analysis agent for a zinc producer
Nexa reported $370,000 in yearly savings from two predictive models and a 9x gain in IT contract analysis
Problem
Nexa is a zinc miner and processor with operations in several countries. Its legacy systems and siloed data limited how far it could scale analytics and AI.
Approach
Nexa built predictive models for zinc filtration and cadmium removal on the Databricks platform. It also built an AI agent that analyzes IT supplier contracts, using pretrained models such as GPT-4 and Llama 3.1 managed through MLflow.
Results
Takeaway
Nexa gave a dollar figure only for the two process models together and a multiple for the contract agent. The company did not say how either was measured.
Categories
- Industry
- Chemicals and materials
- Business function
- Production and qualityFinance, risk and legal
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and MLAI agents
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Vendor case study
Sources
Source summaries
Databricks describes how Nexa used predictive models for zinc filtration and cadmium removal and an AI agent for IT supplier contract analysis. It reports $370,000 in annual savings from the two models and a 9x gain in contract processing efficiency. The page does not give baselines or explain how the figures were calculated.
Full details are in the original sources.
