Custom Software Engineer

Project Role : Custom Software Engineer
Project Role Description : Develop custom software solutions to design, code, and enhance components across systems or applications. Use modern frameworks and agile practices to deliver scalable, high-performing solutions tailored to specific business needs.
Must have skills : AI Agents & Workflow Integration
Good to have skills : Virtual Agents, Generative AI
Minimum 7.5 year(s) of experience is required
Educational Qualification : 15 years full time education

Summary:
As a Custom Software Engineer, a typical day involves designing, coding, and improving various components within systems or applications to meet unique business requirements. The role includes working with modern development frameworks and following agile methodologies to ensure the delivery of scalable and efficient software solutions. Collaboration with different teams and continuous enhancement of software components are integral parts of the daily workflow, focusing on creating tailored solutions that align with organizational goals and client expectations.

Roles & Responsibilities:
- Expected to be an SME, collaborate and manage the team to perform.
- Responsible for team decisions.
- Engage with multiple teams and contribute on key decisions.
- Provide solutions to problems for their immediate team and across multiple teams.
- Lead the implementation of best practices to improve software development processes and team productivity.
- Mentor junior team members to support their professional growth and skill development.
- Coordinate cross-functional efforts to ensure alignment of project objectives and timely delivery.
- Design and build the Agent Orchestration Service for Akura.
- Integrate AI/rule-based agents with Kafka-based event flows.
- Consume events from Kafka topics such as: guidewire.claim.event.created, incident.evaluation.requested, agent.incident.decision.created, incident.creation.requested, audit.event.recorded
- Invoke the appropriate AI agent, rule engine, or decision workflow based on the event type.
- Publish structured agent decision events back to Kafka for downstream services.
- Ensure agent decisions are consumed by ServiceNow Connector, Audit Service, and APEX UI.
- Build or integrate the ServiceNow incident triage agent for Akura.
- Evaluate Guidewire claim/system events and determine whether an incident should be created, updated, ignored, or routed for human review.
- Generate structured recommendations including: Incident required / not required, Severity, Assignment group, Probable classification, Duplicate incident indicator, Recommended next action, Human approval requirement
- Support incident triage workflows aligned to ServiceNow integration needs.
- Build context enrichment logic before invoking the agent.
- Retrieve relevant context from: Guidewire APIs/events, PostgreSQL platform data, ServiceNow incident history, Runbooks and known-error library, Audit/event history, Observability logs and alerts
- Prepare structured context packets for the agent.
- Implement Retrieval-Augmented Generation patterns where knowledge base or runbook context is required.
- Ensure the agent does not rely only on raw Kafka payloads for decisions.
- Define and implement structured agent decision payloads.
- Ensure all agent outputs are machine-readable and auditable.
- Include mandatory metadata such as: Event ID, Correlation ID, Agent name and version, Source event reference, Decision, Confidence score, Reason summary, Recommended action, Human approval flag, Timestamp
- Work with the Schema Governance Engineer to align decision events with schema standards.
- Ensure downstream services do not parse free-form AI text for critical decisions.
- Implement guardrails around agent behavior and tool execution.
- Ensure agent decisions follow defined business rules, approval policies, and risk controls.
- Support decision modes such as: Auto-create incident, Recommend for user approval, No action / audit only
- Build human-in-the-loop approval routing for medium-risk or low-confidence decisions.
- Prevent direct uncontrolled agent execution against enterprise systems.
- Ensure action execution is routed through governed services such as ServiceNow Connector rather than direct unmanaged agent calls.
- Integrate the agent with approved tool APIs and backend services.
- Support to…