What a College Intern - AI Knowledge Management does at HP:

Knowledge store foundation: inventory, clean, de-duplicate, and organize existing R&D knowledge (specs, minutes, technical notes, experiment learnings).Metadata & labeling: apply a lightweight taxonomy (topic, component, program, owner, revision/date, status) and maintain an acronym/jargon glossary to improve retrieval.RAG preprocessing: convert documents into retrieval-ready formats; perform chunking and prepare content for vectorization/embedding and hybrid search (vector + keyword/BM25).Pilot support: help set up agents to demonstrating retrieval effectiveness on unstructured data using Microsoft Copilot StudioEvaluation: Assess retrieval latency and precision as success metricsIngestion playbook: document learnings and create repeatable steps to ingest new knowledge (latest experiment findings, minutes, updated docs) consistently.
 

What you’ll learn / impact

Build a scalable knowledge store designed for integration with a multi-agent workflow (Orchestrator/Router/Retrieval/Structured Data/Jira/Timeline/Safety/Evaluator).Gain hands-on experience with RAG fundamentals: data cleaning → labeling → chunking → vectorization → indexing, and measuring retrieval quality.
 

Individuals who do well in this role at HP, usually possess:

Currently pursuing a Bachelor’s in Computer Science / Data Science / Artificial Intelligence.Able to commit for a full-time 3 months or longer internship, starting around May 2026.Some experience with data preprocessing (Python preferred) and handling structured/unstructured files (CSV, Excel, text/PDF).Basic understanding of semantic search or RAG workflows (coursework, project, hackathon, or self-learning acceptable).Comfortable working with documentation, spreadsheets, and organizing technical content.Ability to follow HP security, confidentiality, and approved AI usage guideline for data handling.Good to have: Familiarity with embeddings/vector databases, BM25/keyword search concepts, or LLM toolchains. Exposure to Microsoft Copilot Studio / SharePoint / OneDrive content management workflows.

Expected deliverables by end of internship

Cleaned + labeled pilot knowledge corpus with a documented taxonomy.Reusable ingestion checklist/playbook for new knowledge updates.Pilot retrieval report (example queries, citation quality, common failure modes, improvement actions).

Sustainable impact is HP’s commitment to create positive, lasting change for the planet, its people, and our communities. This serves as a guiding principle for delivering on our corporate vision – to create technology that makes life better for everyone, everywhere.

HP is a Human Capital Partner – we commit to human capital development and adopting progressive workplace practices in Singapore.

 

#Li-Post

We are looking for an undergraduate intern to support an R&D knowledge retrieval initiative by cleaning, labeling, and preparing existing technical knowledge (documents + datasets) to power a Retrieval‑Augmented Generation (RAG) knowledge store for an agentic multi‑agent system.