Intermediate AI Engineer - LLM Agents, Python

About Shelvz 

At Shelvz, we specialize in developing an advanced platform tailored for the Consumer-Packaged Goods (CPG) industry. Our platform offers a holistic solution for businesses to efficiently gather, consolidate, and analyze data from various sources. This enables our clients to make informed, data-driven decisions, optimize their operations, and achieve a competitive advantage. Our ambition is to become the leading partner for CPG businesses, helping them unlock the full potential.

We are looking for a skilled AI Engineer to join our dynamic team at Shelvz. You should have hands-on experience building applications on top of Large Language Models (LLMs), strong Python skills, and a solid understanding of SQL and relational databases. You will design, build, and improve our AI-powered data analyst. It is a multi-tenant agent that answers business questions in plain language by querying each client's data warehouse, generating charts, and returning clear insights. Experience with Java/Spring is a plus, as it will help you integrate AI features with our core platform.

Responsibilities:

• Design, develop, and maintain LLM-powered agents and services in Python, including tool-use loops, text-to-SQL, and data-visualization tools.
• Write prompts and system instructions, test them, and improve them over time so that answers are accurate, reliable, and grounded in real data.
• Build and maintain integrations between AI services and our platform, including REST APIs, webhooks, and Model Context Protocol (MCP) servers.
• Optimize AI features for cost, latency, and scalability using techniques such as prompt caching, context management, and model selection.
• Implement monitoring, logging, and usage/cost tracking, and troubleshoot issues in production.
• Work with data engineering on ETL pipelines and data warehouse schemas so the AI can understand and query our clients' data.
• Make sure AI features are secure and safe across tenants: read-only data access, authentication, and protection against prompt injection.
• Collaborate with cross-functional teams to define, design, and ship new AI features.
• Participate in code reviews and provide constructive feedback to peers.
• Stay updated with the latest developments in LLMs, agent frameworks, and AI tooling.

Requirements:

• Bachelor's degree in Computer Science, Engineering, Data Science, or a related field.
• 2–4 years of software development experience, with strong Python skills.
• Hands-on experience building applications with LLM APIs (e.g. Anthropic Claude, OpenAI), including function calling / tool use.
• Strong SQL skills and a solid understanding of PostgreSQL or other relational database management systems.
• Experience building backend services and REST APIs (Flask, FastAPI, or similar).
• Understanding of prompt engineering, context-window management, and how to evaluate LLM output quality.
• Familiarity with Git and modern development workflows.
• Strong problem-solving skills and attention to detail.
• Excellent communication and collaboration abilities.

Nice to Have:

• Experience with agentic frameworks or protocols such as the Model Context Protocol (MCP) or the Claude Agent SDK.
• Experience with RAG, embeddings, or vector databases.
• Knowledge of data warehousing, ETL pipelines, and BI/analytics.
• Experience with data visualization libraries (e.g. Matplotlib, Plotly).
• Knowledge of Java and the Spring framework.
• Familiarity with cloud services, Docker, and CI/CD.
• Experience building multilingual applications, especially Arabic.
• Knowledge of Agile methodologies.
• Experience in the CPG, retail, or field-force industries.