We are designing and building a Category Intelligence Engine (CIE) on the client's AWS workplace. The platform combines a governed procurement data foundation with an LLM-based intelligence layer that converts purchase order, contract, specification and market data into strategic recommendations. The engineer is a core member of our Data and Intelligence Layer Enablement workstream and will work closely with Kearney consultants, client functional and tech stakeholders.
Scope of the role
- Build and harden the intelligence layer on Amazon Bedrock, build 3 interconnected agents
- Implement retrieval over procurement documents (specifications, contracts, industry price circulars, tender data) using hybrid search (keyword and vector retrieval),
with appropriate chunking and re-ranking
- Establish the agent harness: tool orchestration, structured outputs, guardrails, evaluation sets and regression testing, ensuring outputs are reproducible and defensible with the client
- Develop document extraction pipelines (Textract / Bedrock) for specification and contract PDFs
- Build backend services (Lambda, API Gateway, Cognito) supporting the React front end
Request your help in getting a couple of solid FDEs for quick evaluation and deployment.