Job Summary
Data Science Skills & Work Order Details :
Skills required:
Experience with Google ADK is an important requirement for this role; however, we are seeking a broader skill set that includes applied machine learning, large language models (LLMs), agentic AI, prompt engineering, experimentation, model evaluation, and the practical delivery of AI solutions into production environments.
Scope Summary:
Design, build and operationalise a fully automated Best Claim Handler (BCH) decisioning capability within the Claims Automation project that consumes claim assessment outputs, applies routing and assignment logic, and returns the best claim handling recommendation to ensure claims are allocated to the most appropriate claim handler, specialist team or processing route.
The solution must:
- Consume fraud propensity scores, iapv outcomes, claim authentication outputs and claim attributes required for routing decisions.
- Apply business rules and AI-driven decisioning logic to determine the routing recommendation.
- Generate routing decisions in near real time to support FNOL and downstream claims handling processes.
- Support assignment to claim handlers, specialist teams and straight-through processing pathways.
- Apply consistent and explainable routing decisions aligned to operational and customer outcomes.
- Detect, monitor and report routing quality, service performance and decisioning exceptions.
- Minimise manual intervention through automated orchestration, monitoring and recovery processes.
- Ensure routing decisions are auditable and reusable within the broader Claims Automation platform.
High-level Delivery Overview:
1. Consume & Standardise Routing Inputs
- Establish automated consumption of BCH input data.
- Consume FNOL Fraud Model scores, IAPV outputs and Claim Authenticator outputs.
- Apply common business definitions and standards.
- Validate completeness, conformity and data quality.
2. Assess Claim Characteristics
- Evaluate claim complexity and handling requirements.
- Analyse fraud signals, authentication outcomes and claim attributes.
- Derive routing attributes required for decision making.
- Generate reusable claim routing features.
3. Curate Best Claim Handler Decisions
- Apply business rules and decisioning logic.
- Determine the optimal claim handling route.
- Create Specialist Team and Straight-Through Processing Recommendations.
- Ensure routing decisions are aligned to customer outcomes, operational efficiency and claim resolution objectives.
4. Integrate Routing Decisions
- Integrate BCH outputs into the Claims Automation workflow.
- Return routing decisions to Pega and downstream claims platforms.
- Support automated and assisted claims handling journeys.
- Maintain full auditability of routing decisions.
5. Automated Refresh & Operationalise
- Implement fully automated orchestration of BCH decisioning services.
- Monitor routing effectiveness, decision quality and service health.
- Track business outcomes and operational performance metrics.
- Minimise manual intervention through automated controls, alerting and recovery processes.
Key Responsibilities
Data Science Skills & Work Order Details :
Skills required:
Experience with Google ADK is an important requirement for this role; however, we are seeking a broader skill set that includes applied machine learning, large language models (LLMs), agentic AI, prompt engineering, experimentation, model evaluation, and the practical delivery of AI solutions into production environments.
Scope Summary:
Design, build and operationalise a fully automated Best Claim Handler (BCH) decisioning capability within the Claims Automation project that consumes claim assessment outputs, applies routing and assignment logic, and returns the best claim handling recommendation to ensure claims are allocated to the most appropriate claim handler, specialist team or processing route.
The solution must:
- Consume fraud propensity scores, iapv outcomes, claim authentication outputs and claim attributes required for routing decisions.
- Apply business rules and AI-driven decisioning logic to determine the routing recommendation.
- Generate routing decisions in near real time to support FNOL and downstream claims handling processes.
- Support assignment to claim handlers, specialist teams and straight-through processing pathways.
- Apply consistent and explainable routing decisions aligned to operational and customer outcomes.
- Detect, monitor and report routing quality, service performance and decisioning exceptions.
- Minimise manual intervention through automated orchestration, monitoring and recovery processes.
- Ensure routing decisions are auditable and reusable within the broader Claims Automation platform.
High-level Delivery Overview:
1. Consume & Standardise Routing Inputs
- Establish automated consumption of BCH input data.
- Consume FNOL Fraud Model scores, IAPV outputs and Claim Authenticator outputs.
- Apply common business definitions and standards.
- Validate completeness, conformity and data quality.
2. Assess Claim Characteristics
- Evaluate claim complexity and handling requirements.
- Analyse fraud signals, authentication outcomes and claim attributes.
- Derive routing attributes required for decision making.
- Generate reusable claim routing features.
3. Curate Best Claim Handler Decisions
- Apply business rules and decisioning logic.
- Determine the optimal claim handling route.
- Create Specialist Team and Straight-Through Processing Recommendations.
- Ensure routing decisions are aligned to customer outcomes, operational efficiency and claim resolution objectives.
4. Integrate Routing Decisions
- Integrate BCH outputs into the Claims Automation workflow.
- Return routing decisions to Pega and downstream claims platforms.
- Support automated and assisted claims handling journeys.
- Maintain full auditability of routing decisions.
5. Automated Refresh & Operationalise
- Implement fully automated orchestration of BCH decisioning services.
- Monitor routing effectiveness, decision quality and service health.
- Track business outcomes and operational performance metrics.
- Minimise manual intervention through automated controls, alerting and recovery processes.