Early Careers: Intern, AI Prompt Engineer

Aon Corporation

23h•$26•Hybrid

About The Position

The Prompt Engineer Intern will support the design and implementation of AI-powered solutions focused on document intake and intelligent agent development. This role will work closely with AI engineers, software engineers, and product stakeholders to build and refine prompts, retrieval pipelines, and document Q&A/assistant workflows. The intern will spend most of their time working with internal documents (PDFs, emails, reports) to improve how information is ingested, organized, and surfaced through large language models. Opportunities are availaible in: Chicago, IL Bannockburn, IL Aon’s Internship Program Interns within Aon will gain valuable experience assisting our teams supporting both client and internal projects. Interns will work with a cross-functional team to gain visibility and experience across a broad segment of clients. The 10-week summer program includes weekly spotlights delivered by Aon business leaders, a community outreach event and networking opportunities with fellow interns and colleagues.

Requirements

Currently pursuing a degree in Computer Science, Data Science, Mathematics/Statistics, Cognitive Science, Information Science, or a related technical field.
Familiarity with Python for scripting, data processing, and prototyping.
Basic understanding of how large language models (LLMs) work, including their strengths, limitations, and common failure modes.
Prior exposure to at least one LLM platform or API (e.g., experience calling LLMs programmatically, using SDKs, or working with hosted LLM tools).
Ability to read technical documentation and quickly translate it into working prototypes.
Clear written communication skills, especially for documenting experiments and summarizing findings.
Demonstrated curiosity, willingness to learn new tools quickly, and comfort working in an iterative, experiment-driven environment.
Ability to work effectively with a global, cross-functional team and respect diverse perspectives, communication styles, and ways of working.

Nice To Haves

Experience working with unstructured text data such as PDFs, emails, or reports.
Familiarity with concepts related to retrieval-augmented generation (RAG), document chunking, and embeddings.
Experience designing or running evaluations/experiments (e.g., A/B tests, simple metric dashboards, or offline test sets).
Exposure to building or using agents/workflows that call external tools/APIs.
Understanding of basic data quality, security, and privacy considerations when handling internal or sensitive documents.

Responsibilities

Build and maintain document ingestion scripts and pipelines for PDFs, emails, and reports.
Design and refine metadata schemas and taxonomies to improve document organization and retrieval.
Clean, normalize, and structure unstructured text to support downstream LLM use cases.
Contribute to the development and improvement of retrieval-augmented generation (RAG) systems.
Evaluate retrieval quality (e.g., coverage, recall/precision) and propose improvements.
Apply security and privacy awareness when handling documents, including PII/redaction workflows.
Help design and refine LLM-powered agents such as document Q&A assistants, research/summary agents, and workflow assistants that call tools/APIs.
Configure and iterate system prompts, instructions, and behavior guidelines for agents.
Implement and maintain simple workflows that connect agents to existing tools and APIs.
Log, investigate, and analyze agent errors or hallucinations, and suggest prompt or configuration changes.
Design and test prompts for document understanding, summarization, and Q&A use cases.
Create and maintain evaluation datasets and test cases for prompts and agents.
Run prompt and agent experiments (including simple A/B tests) and quantify improvements using basic metrics.
Document findings and best practices for prompt patterns, guardrails, and failure modes.
Work on scoped, independent mini projects with regular mentorship and feedback.
Collaborate with product owners, scrum masters, and domain experts to clarify requirements and translate them into concrete technical tasks.
Contribute to light documentation (e.g., experiment notes, runbooks, internal guides) to support knowledge sharing across the team.