Principal AI Engineer

Purpose

Principal Software Development Engineers own a company-level technical problem space on a multi-quarter horizon. In this role the problem space is AI: how Arrivy's platform predicts, recommends and acts for the dispatchers, field crews and customers who rely on it.

Arrivy already ships AI inside the product. Scheduling suggestions rank the right crew for a job, routes are optimised, field photos are scored against reference images, and completed jobs are summarised automatically. As our most senior AI engineer, you set the direction for what comes next and make the architectural calls the rest of engineering builds on.

What you'll own

• Own the technical strategy for AI across the platform: which problems we solve with models, which with rules, and which we leave alone.
• Make the architectural calls that are expensive to reverse, and document the reasoning: build versus buy on models, how predictions and agents reach into scheduling and dispatch, where customer data may and may not go.
• Own the evaluation standard: how we know an AI feature is right before it ships, and how we notice when it stops being right in production.
• Design the guardrails for AI that takes actions. A reschedule triggers a customer notification, so a wrong action reaches a real customer. You decide what needs human confirmation and how user permissions carry through.
• Own the cost and latency envelope of AI features, and advise the leadership team on technical risk, cost and feasibility in terms they can act on.
• Build hands-on where it matters most: the first version of the hardest piece, and the reference implementation others copy.
• Raise the ceiling on engineering quality across teams, and mentor Staff and Senior engineers toward the same scope.

What we're looking for

• 10+ years of software engineering, including sustained impact at Staff level or equivalent, with several of those years shipping machine learning or AI features in a production product.
• Demonstrated ownership of platform-level architecture in a production SaaS business.
• Depth across the AI stack you will be deciding on: language models and agents (tool use, retrieval, evaluation) and classical machine learning for prediction or optimisation. You need not be expert in both, but you can judge both.
• Working knowledge of the main model families and what separates them in practice: OpenAI's GPT, Anthropic's Claude, Google's Gemini, and open-weight models such as Llama and Mistral. You can compare them on quality, cost, latency and data terms, and design so we can switch between them.
• Experience with the platforms these models run on, such as Amazon Bedrock, Microsoft Foundry and Google's Gemini Enterprise Agent Platform (formerly Vertex AI), as well as direct provider APIs. You understand their security, data residency and pricing models.
• A record of taking AI from prototype to production, including evaluation, monitoring, cost control, failure handling and rollback.
• Strong production engineering in Python or a comparable backend language. You still write code.
• Ability to hold a technical position under commercial pressure, and to change it on evidence.
• Exceptional written communication in English and the patience to build agreement slowly, including with colleagues in other time zones.
• Commercial literacy: you understand what the technical choice costs the business.

Bonus points

• Depth in field service, logistics, scheduling or routing, including optimisation methods such as vehicle routing and constraint solving.
• Experience with computer vision in production.
• Experience opening a product to AI assistants through agent protocols such as the Model Context Protocol.
• Experience helping to establish a new engineering site.
• A track record of developing other senior engineers.

Pay: QAR18,000.00 - QAR23,000.00 per month

Work Location: In person