AI Agent Operational Lift for Harold B. Lee Library in Provo, Utah

What the Harold B. Lee Library Does

The Harold B. Lee Library (HBLL) at Brigham Young University is a major academic and research library serving over 30,000 students and faculty. Founded in 1882, it houses millions of volumes, extensive special collections, and vast digital archives. Its core mission is to acquire, preserve, and provide access to information resources that support the university's teaching, learning, and research objectives. Operating with a staff of 501-1000, it functions as both a traditional library and a modern digital gateway, managing physical collections, digital repositories, research support services, and instructional programs.

Why AI Matters at This Scale

For a large academic library like HBLL, AI is not a luxury but a strategic necessity to manage scale and complexity. The sheer volume of digital assets and the high demand for efficient research support from a large user base create a perfect storm that manual processes cannot address. AI offers the tools to move from being a reactive repository to a proactive research partner. It can unlock the latent value in millions of digitized items, personalize the daunting research journey for students, and allow a sizable but finite staff to focus on high-touch, expert services rather than repetitive tasks. At this institutional scale, even marginal efficiency gains in discovery or curation translate into significant time and resource savings, directly supporting the university's academic mission.

Concrete AI Opportunities with ROI Framing

1. AI-Powered Universal Search Engine: Implementing a neural search layer over all digital collections—catalogs, databases, institutional repositories, and archival finding aids—can drastically reduce the time users spend finding relevant materials. ROI is measured in increased resource utilization, higher user satisfaction, and more productive research output, justifying the integration cost through amplified academic impact.

2. Automated Metadata Generation for Archives: Applying computer vision and NLP to digitized special collections (photos, letters, manuscripts) can auto-generate descriptive tags, transcripts, and summaries. The ROI is clear: it would take decades of staff time to manually process backlogs. AI accelerates public access to heritage materials, fulfilling preservation mandates and attracting research interest.

3. Predictive Collection Development: Using AI to analyze citation trends, interlibrary loan requests, and academic publication forecasts allows for data-driven acquisition decisions. ROI is realized through optimized budget allocation, ensuring funds are spent on materials with the highest future scholarly demand, reducing wasteful spending on low-use items.

Deployment Risks Specific to This Size Band

Libraries of this size (501-1000 employees) face unique AI adoption risks. Integration Complexity: They typically operate a patchwork of legacy integrated library systems (ILS), digital asset managers, and vendor databases. Integrating AI tools without disrupting core services requires careful API management and middleware, posing a significant technical hurdle. Budgetary Constraints: As part of a public university, capital for experimental technology is often limited and competed for. AI projects must demonstrate clear, often non-financial, mission-aligned ROI to secure funding. Change Management: A large, established staff with deep expertise in traditional librarianship may view AI as a threat rather than a tool, risking low adoption. Successful deployment requires inclusive training and framing AI as augmenting, not replacing, professional judgment. Data Governance & Bias: Using patron data for personalization raises privacy concerns. Furthermore, AI models trained on historical collections could perpetuate archival biases, requiring robust governance frameworks to ensure ethical, equitable outcomes.