About this role

AI application development: Build and maintain Python components that connect applications to approved AI models, internal services, and databases.
LLM integration: Assist with prompt development, model API integration, structured outputs, and validation of model responses.
Retrieval workflows: Support document preparation, chunking, embedding generation, and retrieval-augmented generation (RAG) pipelines.
Orchestration: Help connect model calls, retrieval steps, tools, and business logic into controlled workflows with appropriate error handling and human review.
Evaluation: Create test cases and assess output relevance, factual consistency, response time, and common failure scenarios.
Application support: Investigate failed API calls, retrieval problems, unexpected model responses, and workflow errors; document and escalate findings.
Responsible implementation: Use approved AI coding assistants, validate generated code, and follow security and data-handling requirements. Do not submit confidential information to unapproved AI services.

0–2 years of relevant experience; a degree in computer science, engineering, data science, or a related field, or equivalent practical experience.
Python and APIs: Basic proficiency in Python, HTTP/REST APIs, JSON, exception handling, and integrating external services.
Databases: Basic SQL and relational-database concepts, plus awareness of the purpose of vector search and embeddings.
Machine learning: Understanding of core ML concepts, including training versus inference, evaluation, and overfitting.
LLMs: Foundational knowledge of prompting, tokens, context windows, embeddings, RAG, hallucinations, and output validation.
Orchestration: Basic understanding of multi-step AI workflows, tool calls, state, retries, and fallbacks; exposure to one tool or framework such as LangChain, LangGraph, n8n, or an equivalent.
AI-assisted development: Ability to use a coding assistant responsibly, explain generated code, and verify its behavior through testing.
Engineering and support: Basic Git, debugging, testing, log analysis, and awareness of secure credential handling.
Preferred qualifications: Practical AI work: A chatbot, document Q&A, retrieval, or AI automation project built during study, an internship, or independent learning.
Application development: Exposure to FastAPI or Flask, Docker, a cloud platform, or an approved model-serving service.
AI safety and monitoring: Awareness of prompt injection, sensitive-data exposure, output guardrails, and monitoring model usage and errors.

AI Engineer