Artificial Intelligence Engineer

Role Overview

AI Engineers are needed to execute the day-to-day implementation of the agentic platform across

markets, working under the technical direction of Senior AI Engineers and AI Lead. You will build

individual agents, implement tool-use integrations, write prompts, and run evaluation experiments.

This is a high-execution role requiring strong Python skills, solid LLM knowledge, and the ability to

deliver features independently with clear guidance from seniors.

Core Responsibilities

• Implement individual agents and sub-agents within the ADK framework including intent routing

agents, FAQ agents, transactional agents, and escalation agents.

• Write, test, and iterate on system prompts templates for each market use case.
• Build tool-use integrations connecting agents to external APIs: knowledge bases, MCP lookup,

booking systems, and notification services.

• Run offline and online evaluation experiments using standard chatbot metrics (intent accuracy, task

completion, CSAT) and document findings.

• Participate actively in code reviews; implement feedback from Senior AI Engineers/AI Lead promptly.
• Maintain vector database collections: chunking strategies, embedding models, indexing pipelines,

and retrieval evaluation.

• Support AI Translators in localizing prompts and agent responses for each market language —

providing technical input on character limits, tokenization quirks, and language-model behaviour.

• Reproduce, document, and escalate production bugs related to agent behaviour, model

hallucinations, and tool failures.

• Keep implementation documentation up to date: agent flow diagrams, prompt registries, and test

Must-Have Technical Skills

Python (Proficient) Google ADK basics Prompt engineering LLM APIs (Gemini / OpenAI) RAG pipelines Vector DBs Git REST API consumption Data analysis GCP basics JSON / YAML Pytest

Key Qualifications

• 5+ years of AI/ML engineering experience; at least 2 year building or integrating conversational AI or

LLM-based products.

• Demonstrable hands-on experience with Google Cloud AI stack
• Solid understanding of transformer models, tokenisation, context windows, and prompt sensitivity.
• Familiarity with agentic patterns: ReAct, plan-and-execute, reflection loops, and multi-agent handoff

protocols.

• Comfortable working in an agile sprint environment with clearly defined tickets and daily standups.
• Basic understanding of responsible AI and data privacy considerations — knowing when to flag

concerns to seniors.