Prompt Engineer



This role will provide a vital role in our GenAI team, responsible for developing and optimizing prompts to enhance the performance and output quality of our AI models. The role responsibilities include: The role requires critical thinking and a high degree of collaboration, and business understanding as well as attention to detail. The characteristics include being versatile, displaying leadership qualities and enthusiasm to take on new problems, ultimately assisting in moving technology forward. This role will have varying degrees of analysis, design, development, documentation, testing and support responsibilities.

Essential capabilities:

  • Enthusiasm for technology, keeping up with latest trends
  • Ability to articulate complex technical issues and desired outcomes of system enhancements
  • Proven analytical skills and evidence-based decision making
  • Excellent problem solving, troubleshooting & documentation skills
  • Strong written and verbal communication skills
  • Excellent collaboration and interpersonal skills
  • Strong delivery focus with an active approach to quality and auditability
  • Ability to work under pressure and excel within a fast-paced environment
  • Ability to self-manage tasks

Desired Experience:

  • 6-10 Years technical experience
  • Experience in written communication of business concepts
  • Understanding of ethical considerations and bias mitigation techniques in AI model development
  • Experience with Langchain
  • Background in statistical analysis and experimental design for evaluating prompt effectiveness and model performance.
  • Experience with AWS, Azure, and Databricks for scalable and efficient AI model training and deployment
  • Familiarity with AI model development and deployment workflows
  • Experience with CI/CD pipeline deployments
  • Strong programming experience in Python and data manipulation with use of appropriate Python packages
  • Hand-on experience with NLP techniques and technologies

To apply for this job please visit

Share this :