Use cases

From process challenges
to practical AI applications.

Explore how BFAI applies causal analysis, visual inspection AI, MLOps and manufacturing knowledge across semiconductor production.

Foundry

Predict process outcomes before a run

Challenge: Process outcomes can be difficult to anticipate before a wafer enters production equipment.

Approach: Combine equipment history and sensor data with causal analysis to predict process outcomes and support parameter optimization.

Product Causal AI Platform
Foundry

Investigate the causes of yield loss

Challenge: Yield loss can involve multiple processes and tools, making root-cause analysis complex and time-consuming.

Approach: Connect process, equipment and inspection data through causal graphs, explore counterfactual scenarios and explain findings with a local LLM.

Products Causal AI Platform · Ontology-based Local LLM
IDM

Bring manufacturing insights back to design

Challenge: Valuable manufacturing insights do not always reach design teams or inform future product development.

Approach: Structure causal findings and process knowledge, share relevant insights with design teams and support future design improvements.

Products Causal AI Platform · Ontology-based Local LLM
IDM

Automate visual inspection and maintain accuracy

Challenge: Manual defect review creates workload, while changing inspection conditions can affect model performance.

Approach: Deploy visual inspection AI on edge PCs, monitor performance and manage model retraining and updates through MLOps.

Products AI Models · Enterprise MLOps
IC package substrates

End-to-end substrate inspection automation

Challenge: Automated optical inspection still leaves substantial manual verification and defect classification work.

Approach: Apply AI-assisted verification and classification to existing inspection workflows, integrating model operations with site quality processes.

Products AI Models · Enterprise MLOps
Equipment makers

Bring expert knowledge to every field engineer

Challenge: Technical documents, service records and troubleshooting experience are difficult to retrieve and reuse.

Approach: Structure equipment knowledge and retrieve relevant technical information and guidance using an on-premises local LLM.

Product Ontology-based Local LLM

Start with your manufacturing challenge.

Talk to us about your process, data and deployment environment.

TOP