Senior AI Prompt Architect

MCI Careers; Wes-Kaap; Hybrid; ZAR 1,959,823 - 2,939,734; Full time

Job summary

MCI Careers is seeking a Senior AI Prompt Engineer to design, optimize, and manage prompts that maximize the effectiveness of Large Language Models and AI-powered applications. You will work with AI engineers, product managers, and stakeholders to develop scalable prompting solutions that drive measurable business outcomes.

You will design prompts for conversational AI and automation, test and refine prompts, and contribute to AI performance evaluation and best practices across multiple use

Qualifications

• Bachelor’s degree in a relevant field such as CS, AI, linguistics, communications or information systems.
• Minimum 5 years in AI, software development, content strategy, research, analytics, or related discipline.
• Minimum 2 years hands‑on with Generative AI technologies and prompting for LLMs.

Responsibilities

• Design and maintain prompts to enhance AI system performance.
• Develop reusable prompt templates for conversational AI and business apps.
• Test prompts and conduct A/B evaluations to improve outcomes.
• Analyze AI outputs and identify optimization opportunities.
• Collaborate with AI engineers, product managers, trainers and stakeholders.
• Translate business requirements into prompting strategies and support implementations.

Skills

Prompt design

Prompt testing

AI evaluation

Collaboration

Research

Communication

Education

Bachelor's degree in CS/AI/Linguistics

Tools

OpenAI

Anthropic

Gemini

Claude

Job description

MCI Careers is seeking a Senior AI Prompt Engineer to design, optimize, and manage prompts that maximize the effectiveness of Large Language Models and AI-powered applications. You will work with AI engineers, product managers, and stakeholders to develop scalable prompting solutions that drive measurable business outcomes.

You will design prompts for conversational AI and automation, test and refine prompts, and contribute to AI performance evaluation and best practices across multiple use