Halliburton enhances seismic workflow creation with Amazon Bedrock and Generative AI
Natural language replaces 100-tool manual config in Halliburton's Seismic Engine.
Halliburton's Seismic Engine, a cloud-native application for seismic data processing, previously required manual configuration of approximately 100 specialized tools to create workflows—a time-consuming, error-prone process demanding deep expertise. To address this, Halliburton collaborated with the AWS Generative AI Innovation Center to build an AI-powered assistant. The solution leverages Amazon Bedrock, Bedrock Knowledge Bases, Amazon Nova Lite, and Amazon DynamoDB to transform complex workflow creation into natural language conversations. Geoscientists and data scientists can now configure processing tools by simply describing what they need.
The system uses a FastAPI application on AWS App Runner with a streaming interface. An intent router powered by Amazon Nova Lite classifies user queries as 'Workflow_Generation', 'QnA', or 'General_Question'. For workflow requests, a generation agent using Anthropic's Claude on Amazon Bedrock creates YAML workflows by selecting from 82 available Seismic Engine tools. The Q&A system retrieves relevant answers from indexed documentation using Bedrock Knowledge Bases with OpenSearch Serverless. Evaluation results show workflow acceleration of up to 95%, and the scalable cloud-native architecture makes advanced geophysical tools more accessible to a broader range of users.
- AI assistant reduces seismic workflow creation time by up to 95% using Amazon Bedrock
- Natural language queries replace manual configuration of ~100 processing tools
- System uses intent routing (Amazon Nova Lite) and workflow generation (Claude via Bedrock) with DynamoDB for context
Why It Matters
Democratizes complex seismic analysis, speeding energy exploration and reducing expert dependency by 95%.