Systems Architect

'Role Summary

The AI Architect is responsible for the overall solution design for the AI Search platform. This includes the

ingestion and retrieval architecture, multilingual behaviour, AI guardrails, and integration with the

existing AWS-hosted CMS and master-data layer. The role acts as the main architectural owner for the

project and works closely with the client, delivery leadership, and engineering teams to keep design

decisions clear and practical.

Key Responsibilities

• Own and refine the overall architecture for the Azure-native AI Search platform, keeping the
  
  

ingestion and search/retrieval tracks clearly separated.

• Define how Azure AI Foundry, Azure AI Search, Azure Data Factory, Azure AI Language, Azure
  
  

Functions or Foundry Agent Service, Azure Front Door, APIM, and Cosmos DB work together.

• Set the integration approach for the AWS-hosted Sitefinity CMS and SDMS, which remain unchanged
  
  

and connect through the IHH SG API Gateway.

• Design the doctor-search flow as a rule-based, deterministic path, separate from semantic and
  
  

vector-based content search, with both feeding into a grounded AI summary layer.

• Define multilingual search behaviour, including query-language results where available, fallback to
  
  

English content when no results are found, and summaries generated in the user's original query

language.

• Put practical AI safety controls in place, including Prompt Shields, Azure AI Content Safety checks, PII
  
  

redaction, and sign-off for clinical system prompts.

• Cover the key non-functional areas: security, PDPA requirements, scalability, concurrency, disaster
  
  

recovery, cost control, and observability.

• Shape the Azure DevOps and CI/CD approach, including Bicep IaC, staged promotion from Dev to
  
  

UAT and Production, blue/green release using index aliases and APIM revision swaps, and rollback

planning.

• Prepare and maintain architecture documents, diagrams, and decision records, and lead review
  
  

discussions with the client, healthcare consultants, and CMS vendor.

• Guide the Tech Lead and engineering team on design choices, trade-offs, and boundaries so delivery
  
  

stays aligned with the architecture.