We are seeking a Full Stack Developer to build the human-facing surfaces of an AI-powered banking automation platform at a leading UAE financial institution. You will design and deliver the HITL reviewer interface, the AgentOps operational dashboard, and internal developer portals — translating complex AI agent state, confidence scores, and exception data into interfaces that empower fast, informed human decisions. Working at the intersection of AI systems and operations teams, your work directly impacts straight-through processing rates and the quality of human review across regulated banking workflows.
Key Responsibilities
HITL reviewer interface: Design and build the Ops reviewer queue interface — document image display, extracted field overlays with per-field confidence indicators, exception displays, and Approve/Edit/Reject workflows optimised for reviewer throughput.
AgentOps dashboard: Build and maintain a real-time operational dashboard consuming Azure Monitor and Log Analytics data — STP rate, HITL rate, rejection rate, document volume, token consumption, latency percentiles, and confidence distribution trends.
System integration: Build communication layers between AI systems and banking applications, with deep-linking, session management, and backend API integration.
Internal tooling and portals: Build and maintain Use Case Manifest authoring tools, Prompt Management System UI, and ground-truth annotation interfaces.
API integration: Integrate frontend applications with AI Gateway REST APIs, APIM-governed LLM endpoints, Azure Monitor query APIs, and workflow job status APIs.
Performance and accessibility: Ensure interfaces meet accessibility standards; optimise for performance on low-bandwidth branch connections; implement graceful error handling for AI pipeline latency.
Design system: Build and maintain a consistent component library across all platform-facing interfaces; align with institutional brand standards.
Required Skills & Qualifications
5+ years full stack development experience with strong React/TypeScript frontend skills.
REST API design and consumption; WebSocket/Server-Sent Events for real-time updates.
State management: Redux, Zustand, or React Query for complex async data flows.
Charting and data visualisation: Recharts, D3, or similar for operational metrics dashboards.
Node.js or Python backend (FastAPI/Express) for Backend for Frontend (BFF) patterns.
Authentication: OAuth2/OIDC, Entra ID B2C for customer-facing channels.
Docker and Kubernetes basics — deploying frontend workloads to AKS.
Bachelor's degree in Computer Science, Software Engineering, or equivalent.
Preferred Attributes
Mobile development experience: React Native or native iOS/Android.
Image rendering and annotation: displaying high-resolution document scans with overlay layers.
Azure Static Web Apps, Azure Front Door, and CDN configuration.
Experience rendering streaming API responses and LLM token streams in chat interfaces.
Familiarity with AI/ML concepts — sufficient to meaningfully display confidence scores and model metrics.
Portfolio of production applications (GitHub profile or deployed products).
Halian Group: With over 28 years of experience, we have come to understand that innovation is the only way to provide agile, practical solutions that transform businesses and careers. Our resourcing and smart services help you to realise tomorrow's potential. Discover the amazing things possible when you bring the right people and the right technologies together. At Halian, we recognise that diversity, equity, and inclusion (DEI) are essential to building high-performing teams for our clients. We are committed to connecting organisations with top talent from all backgrounds, ensuring that every individual feels valued, respected, and empowered to contribute their unique perspectives. We encourage applications from all qualified candidates, regardless of race, gender, disability, or any other characteristic that makes them unique. #LI-CC1
AI Fullstack Developer in Abu Dhabi, United Arab Emirates
Bereit für den nächsten Schritt?
Melden Sie sich online an - es dauert nur 10 Minuten.