About the Role
We are looking for a Prompt Engineer to craft, optimize, evaluate, and benchmark prompts for enhanced AI performance at Netomi. This role requires expertise in NLP, AI behavior, and customer-specific business rules. You will collaborate with the Customer Success team and data scientists to develop customized, effective AI solutions.
Responsibilities
-
Prompt Development: Design and refine client-specific prompts, ensuring accuracy and relevance. Define tool descriptions for agentic frameworks to enhance AI interactions.
-
Optimization & Testing: Improve prompts for clarity and performance, automate testing with scripts, and evaluate LLMs for best-fit solutions.
- Evaluation & Benchmarking: Develop evaluation frameworks and benchmark prompts to establish best practices.
-
Collaboration & Documentation: Work with Customer Success and Data Science teams, maintaining clear documentation on prompt development and optimization.
-
Research & Innovation: Stay updated on NLP advancements, experiment with new prompting strategies, and refine model-specific adaptations.
Requirements
-
Bachelor’s or Master’s degree in Computer Science, Data Science, Linguistics, or a related field.
-
Strong understanding of natural language processing (NLP) and machine learning principles.
-
Experience with AI prompt engineering, including writing, optimizing, and evaluating prompts.
-
Proficiency in programming languages such as Python.
-
Familiarity with AI frameworks and libraries (e.g., TensorFlow, PyTorch, Hugging Face).Excellent analytical and problem-solving skills.
-
Strong written and verbal communication skills.
-
Ability to work collaboratively in a team environment and manage multiple projects simultaneously.
Nice to Have
-
Should have a minimum of 5 years of working in data science
-
Experience with large language models (e.g., GPT-3, GPT-4, Llama3) and their applications.
-
Knowledge of data visualization tools and techniques.
-
Background in linguistics or computational linguistics.
-
Experience with agile development methodologies.