Forward Deployed AI Engineer (Dubai, on-site) at SPsoft

Про вакансію

Forward Deployed AI Engineer (Dubai, on-site) Europe, Ukraine HOT About the role We are looking for a hands-on, self-sufficient AI engineer to join a team embedded with a major aviation organization in Dubai. The role is full-time and on-site in Dubai, UAE, for an initial one-year assignment with possible extension. You will work directly with business stakeholders, understand the operational problem, and own the technical solution from discovery and architecture through implementation, deployment and production operation. This is a dual-skill, engineering-led role rather than a prompt-engineering or voice-only position. The priority is a self-sufficient engineer who can implement and deploy agentic-AI solutions in Python while also configuring a low-code conversational AI platform and refining prompts, dialogue logic and user experience. The candidate does not need to be best-in-class in both disciplines, but must be able to deliver independently beyond prompting alone. What you will do Current platform delivery. Configure, roll out and continuously improve use cases on the current low-code conversational AI platform, including evaluations, drift monitoring, prompt and dialogue refinement, feedback loops and release quality. Agentic AI delivery. Design, build, deploy and operate production-grade action agents and agentic workflows, including internal and operational use cases beyond customer-facing conversations. Omnichannel rollout. Extend customer-service journeys from voice and email to web chat, WhatsApp and WeChat, including service queries and booking requests. APIs and MCP. Connect AI agents to enterprise systems and build the missing last-mile integration layer, including APIs, tools, MCP servers or connectors, webhooks, queues and secure integration patterns. Multi-agent systems. Design orchestrators and clearly scoped specialized agents with reliable handoffs, retries, fallbacks and human review where required. Production quality. Own testing, observabili