Workflow Engineer

🤖 Workflow Engineer — Orkes Conductor ⏳ 6-month initial contract - with extensions📍 Abu DhabiOrchestrate the Workflows That Power Enterprise AI BankingWe’re partnering with one of the UAE’s leading financial institutions to hire a Workflow Engineer (Orkes Conductor Specialist) to design and build the orchestration backbone of a next-generation Agentic AI platform. This is a highly specialised engineering role sitting at the intersection of workflow orchestration, distributed systems, AI agent execution, and core banking integration.You will be responsible for designing how complex banking processes—such as cheque clearing, exception handling, HITL approvals, and compensating transactions—are orchestrated across AI systems, core banking platforms, and human operators. Using Orkes Conductor, you will define the durable workflow layer that binds together LangGraph-based AI agents, ensuring every process is observable, resilient, and compliant with enterprise banking standards.As a Workflow Engineer, you will own the design and implementation of the business process orchestration layer within the AWA platform.You will translate real-world banking processes into durable, distributed workflows, ensuring seamless coordination between AI agents, human reviewers, and core banking systems.You will be the critical bridge between:AI Agent execution (LangGraph / AI Gateway)Core banking systems (ICCS, CBS)Human-in-the-loop operationsEnterprise observability and monitoring systemsWhat You’ll Be DoingOrkes Workflow EngineeringDesign and implement complex workflows using Orkes Conductor.Build task sequences, parallel execution flows, sub-workflows, and SLA-driven orchestration logic.Define workflow structures for all AI banking use cases.Human-in-the-Loop (HITL) OrchestrationConfigure Human Task routing for operational review queues.Implement HITL approval workflows (Approve / Edit / Reject).Ensure seamless resumption of AI workflows after human decisions using LangGraph state restoration.Distributed Systems & Resilience DesignDesign compensating transaction workflows for downstream failures (e.g., core banking or ICCS failures).Implement idempotency patterns using awa_job_id deduplication.Configure retry strategies including exponential backoff, jitter, circuit breakers, and DLQ handling.AI + Workflow IntegrationIntegrate Orkes workflows with LangGraph agent execution models.Manage orchestration of AI tasks as distributed worker processes.Ensure robust error handling between AI and orchestration layers.Observability & MonitoringExport workflow metrics into Azure Monitor / Log Analytics.Build KQL queries to track:Workflow durationTask failure ratesRetry patternsHITL queue depthSLA adherence across use casesWorkflow Governance & DeliveryOwn end-to-end workflow design for new banking AI use cases.Define SLAs and operational expectations with business stakeholders.Design workflow test strategies and validation frameworks.Ensure consistency of workflow standards across the platform.Data Integrity & TraceabilityEnsure awa_job_id propagation across all workflow steps.Validate end-to-end traceability across AI, HITL, and core banking systems.Implement strict schema validation for workflow inputs/outputs.What We’re Looking For - Essential Skills & Experience5+ years’ experience in workflow orchestration or distributed systems engineeringHands-on experience with:Orkes Conductor, Apache Conductor, Temporal or Apache AirflowStrong understanding of distributed systems concepts:IdempotencyEventual consistencySaga patternsAt-least-once delivery semanticsExperience designing long-running, durable workflows with human approval stepsStrong REST API knowledge (especially async and webhook-based patterns)Experience with JSON/YAML-based workflow definitionsPython or Java for workflow worker implementationAdvantageous ExperienceDirect experience with Orkes Conductor (Orkes Cloud / SDK / Human Tasks)Familiarity with LangGraph or LangChain and agentic AI systemsKafka or event-driven architecture (consumer groups, event triggers)Azure Monitor / Log Analytics / KQLBanking domain knowledge (highly beneficial):Straight-through processing (STP)Clearing and settlement workflowsException handling and reconciliation processesWhy Join?This is a rare opportunity to define the orchestration layer of a large-scale enterprise AI banking platform.