Ai Engineer - Alexandria
Summary
The AI Engineer will design and deploy AI-powered features and agents for the Sana Commerce platform, focusing on RAG systems, LLMOps, and building robust evaluation pipelines. The role requires strong software engineering fundamentals and experience integrating AI into distributed, production-grade SaaS applications.
What you'll be doing
- Design, develop, and deliver AI-powered features in Sana Commerce – agents that act on real commerce data (catalog, pricing, inventory, orders)
- Work with modern LLM frameworks and toolchains (e.g. any of Lang*, AWS Bedrock, MS Foundry, Cloudfare Agents, etc…)[VL1.1] to build RAG systems, AI Agents, and dynamic workflows – while understanding the raw agentic loop underneath them
- Work in a cross-functional team (Engineering, Product, Design) to integrate AI seamlessly into a distributed product
- Apply solid AI engineering principles end to end, from context and tool design through evaluation, deployment and continuous improvement
- Build the eval and feedback loops that gate releases: curated datasets, metrics and thresholds that block a regression the way a failing test does
- Design for failure by building the recovery path before the happy path
- Contribute to LLMOps and AI observability, improving reliability, evaluation, and monitoring pipelines
- Stay on top of emerging AI tools and frameworks, helping shape the vision and roadmap of Sana Commerce
What you bring
- Proven commercial experience building and deploying production-grade AI applications.
- 4+ years of web application engineering experience
- Strong software engineering (back-end, front-end nice to have) and SDLC fundamentals (design patterns, SOLID principles, experience with building complex MACH systems, CI/CD, testing)
- Apply strong backend and system design principles to ensure deterministic behaviour where needed, clear failure handling, human-in-the-loop controls, and production readiness.
- Practical knowledge of LLMs, their components (tokenization, attention, embeddings, vector search), and common limitations
- Hands-on experience with RAG pipelines, agents, and Vector Stores and the judgement to know when retrieval is the wrong tool
- Familiarity with LLMOps practices, monitoring, evaluation, and feedback loops.
- Has built real evals that are tied to business impact– dataset, metrics, thresholds and can name the regression they caught
- Cost and latency literacy – can say what a run costs and how to halve it
- Strong working knowledge of Cloud technologies
Ready to build reliability that scales?
Apply now and help shape the foundation of our next-generation SaaS platform.
#LI-JS1
#LI-Hybrid