ML/AI Engineer - Agentic AI & Fraud Analytics (Public Sector)
Core
Design and build agentic AI systems to identify emerging fraud risks and translate them into actionable detection logic for public sector subsidy misuse.
Role type
Senior IC ML/AI Engineer (Agentic AI & Fraud Analytics)
Builds
Production-ready AI workflows, agent systems, and fraud detection logic for public sector training subsidies.
Domain
Public sector / Fraud analytics / Agentic AI
Deliverable
production ML models
Required skills
Python software engineering, LLM application development, agentic AI workflow orchestration, API development, automated testing, AI evaluation frameworks, responsible AI safeguards, enterprise data integration
Preferred skills
AWS services (Bedrock, Redshift, Glue), PySpark, Docker, MLOps tools, RAG, graph-based workflows, fraud detection domain expertise
Technologies
Python, LangGraph, CrewAI, AWS Redshift, AWS Glue, PySpark, Docker, MLflow
Responsibilities
Design end-to-end agentic AI architectures with workflow orchestration and human approval points; Develop AI agents to identify and prioritize evidence-linked fraud risks; Build agent workflows to assess data feasibility and propose risk indicators; Integrate AI workflows with enterprise data platforms (AWS Redshift, Glue, PySpark); Develop modular, tested Python applications and services; Implement automated testing and CI/CD workflows for code and AI components; Establish AI evaluation frameworks for quality, reproducibility, and security; Implement responsible-AI safeguards including access controls and prompt injection protection; Build monitoring dashboards for system health and operating costs.
Seniority
Mid-Senior, hands-on IC
