AddSource is looking for an AI/ML Architect to lead innovative projects in artificial intelligence and machine learning. The role involves working with advanced technologies such as Amazon Bedrock and SageMaker to develop and automate data processing and model fine-tuning for enterprise applications.
Responsibilities:
- Amazon Bedrock (multi-agent orchestration, action groups, guardrails, knowledge bases/RAG)
- Bedrock Data Automation (document parsing for legal PDFs/Word)
- SageMaker (fine-tuning code-specialized LLMs on enterprise COBOL/Java/.NET codebases, embedding models for capability graph, RLHF for context graph self-improvement)
- Prompt engineering (requirement extraction, ambiguity detection, confidence scoring, code generation, test derivation)
- Amazon Textract/Comprehend (legal document structure extraction, entity recognition)
- Neptune/OpenSearch (knowledge graph construction, graph embeddings, cross-mandate pattern detection, BCM taxonomy traversal)
- LangChain/LangGraph or Bedrock multi-agent patterns for the end-to-end forward-pass orchestration from legal language → requirements → impact → code → test