AI Prompt Engineer

Authenticx•Indianapolis, IN

11d•Hybrid

About The Position

The AI Prompt Engineer will lead the design, development, and optimization of large language model (LLM) prompts for scalable conversation scoring at Authenticx. The role translates objective scoring rubrics into reusable, testable prompt logic for enterprise healthcare clients and collaborates with engineering, implementation, product, and data labeling teams to deliver reliable, high-quality prompt workflows. This hands-on role suits someone who thrives in an experimental, iterative environment and is interested in scaling natural language solutions. The role is hybrid and based out of the Indianapolis office. It may include occasional travel for team meetings or planning sessions. Candidates must be authorized to work in the U.S. without current or future visa sponsorship.

Requirements

Strong understanding of LLMs, NLP systems, and prompt engineering concepts.
Ability to structure and document complex logic in a scalable and repeatable manner.
Demonstrated ability to collaborate effectively with technical and non-technical teams.
Analytical mindset and interest in experimentation, iteration, and data-driven decision-making.
Must be authorized to work in the U.S. without current or future visa sponsorship.

Nice To Haves

Experience in structured data labeling or QA workflows.
Familiarity with tools such as LangChain, OpenAI/GPT APIs, or vector databases.
Background in linguistics, cognitive science, or HCI.
Proficiency in Python or scripting tools for evaluation or automation.

Responsibilities

Design and implement sophisticated LLM prompts aligned with internal scoring rubrics and customer-specific needs.
Test and evaluate prompts using labeled datasets to ensure performance accuracy.
Collaborate with engineering to improve tooling and infrastructure for managing LLM scoring workflows.
Maintain a centralized prompt library to promote reuse and standardization across use cases.
Document prompt strategies, testing approaches, and outcomes for internal learning.
Identify and resolve inefficiencies in the prompt development and deployment process.
Partner with product and data labeling teams to improve the quality and consistency of scoring outputs.