Senior AI / Automation Engineer

Client of Salt; Abu Dhabi; On-site; AED 300,000 - 600,000; Full time

Job summary

Client of Salt is seeking a Senior AI/Automation Engineer to design and productionise AI-powered enterprise workflows. You will combine software engineering with hands-on LLM experience, building production-grade AI solutions and tool-calling agents integrated with enterprise APIs.

Responsibilities include document extraction, RAG, context management, guardrails and human-in-the-loop controls, plus monitoring of quality, latency and costs across complex multi-step workflows.

Qualifications

• Strong software engineering background with production AI/LLM experience.
• Advanced Python and API/service development skills.
• Hands-on experience with RAG, embeddings, tool/function calling and structured outputs.
• Experience building asynchronous, stateful and multi-step workflows.
• Strong understanding of testing, observability and production reliability.
• Experience with LLM evaluation, model selection and inference economics.
• Proven track record taking AI solutions from prototype to production.
• Advantage in: Agent orchestration and multi-agent systems, enterprise document processing and OCR, data residency and regulated environments.

Responsibilities

• Build and deploy production LLM and agentic AI workflows.
• Develop tool-calling agents integrated with enterprise APIs and systems.
• Build document extraction, classification and reasoning solutions.
• Implement RAG, context management and structured outputs.
• Integrate AI with deterministic enterprise workflows.
• Build guardrails, validation, confidence thresholds and human-in-the-loop controls.
• Develop automated AI evaluation and regression testing.
• Implement model routing, fallback strategies and inference-cost controls.
• Monitor AI quality, latency, token usage and failure rates.
• Determine when AI should be used versus deterministic software.

Skills

Python

API development

LLM/AI engineering

RAG

Embeddings

Tool calling

Structured outputs

Asynchronous workflows

Observability

Production reliability

Model evaluation

Inference economics

Agent orchestration

Multi-agent systems

Document processing/OCR

Security & data residency

Tools

OCR tools

APIs

Job description

Our client is looking for a Senior AI / Automation Engineer to build and productionise AI-powered enterprise workflows. This is not a chatbot or RAG-focused role . The ideal candidate will combine strong software engineering with hands-on experience in LLMs, agentic AI, document intelligence and enterprise automation . Key Responsibilities: Build and deploy production LLM and agentic AI workflows . Develop tool-calling agents integrated with enterprise APIs and systems. Build document extraction, classification and reasoning solutions. Implement RAG, context management and structured outputs. Integrate AI with deterministic enterprise workflows. Build guardrails, validation, confidence thresholds and human-in-the-loop controls . Develop automated AI evaluation and regression testing. Implement model routing, fallback strategies and inference-cost controls. Monitor AI quality, latency, token usage and failure rates. Determine when AI should be used versus deterministic software.

Desired Candidate Profile

• Strong software engineering background with production AI/LLM experience .
• Advanced Python and API/service development skills.
• Hands-on experience with RAG, embeddings, tool/function calling and structured outputs .
• Experience building asynchronous, stateful and multi-step workflows.
• Strong understanding of testing, observability and production reliability.
• Experience with LLM evaluation, model selection and inference economics .
• Proven track record taking AI solutions from prototype to production .
• Strong Advantage Experience with: Agent orchestration and multi-agent systems Enterprise document processing and OCR AI automation across Finance, Procurement, HR or Legal Multi-model architectures Enterprise security, data residency and regulated environments
• The key mindset: build AI that works reliably in the real world including knowing what to do when the model is wrong