Who owns the resulting work?
The institution does. Workflow maps, decision logs, interview-derived documentation, and playbooks are designed to remain editable and useful after the engagement.
Operational memory · Knowledge infrastructure
Map the work that holds your library together.
LibraryOps is an advisory practice for library leadership. It surfaces the tacit workflows, decisions, and process knowledge that already run your institution — and prepares them, responsibly, for what comes next.
About / The practice
Library experience meets knowledge infrastructure.
LibraryOps is led by an information scientist whose career has centered on the systems that help organizations find, trust, preserve, and use knowledge.
George Jensen · Founder & Principal Consultant
George is based in Rhinebeck, New York, and brings nearly a decade of direct higher-education experience across the Herron School of Art Library and Indiana University / IHETS. He holds a Master of Information Science from Indiana University.
Across more than twenty years, his work has included library reference, metadata and taxonomy, enterprise search, information architecture, data conversion, research repositories, technical delivery, and responsible AI-enabled knowledge systems.
Methods and standards: MARC, OCLC, Dublin Core, ontology, taxonomy, controlled vocabularies, semantic search, workflow analysis, qualitative interviews, and knowledge-system design.
01 / The problem
Institutional memory is walking out the door, one staff member leaving at a time.
Every library runs on workflows that nobody wrote down — passed shoulder-to-shoulder, refined by people who’ve been in the building for twenty years. When leadership turns to AI, modernization, or strategic planning, that hidden infrastructure is what gets in the way. Or, more often, what quietly disappears.
PROBLEM 01
Decisions live in inboxes and in the heads of retiring staff.
The reasoning behind a vendor switch, a discovery layer customization, or a service-point policy is rarely captured anywhere a successor can find it.
PROBLEM 02
Workflows are tacit, taught by demonstration.
The cataloger’s exception-handling. The ILL escalation path. The reference team’s informal triage. None of it shows up in a procedure manual; all of it shapes patron experience.
PROBLEM 03
AI initiatives stall before they start.
You can’t responsibly automate what you can’t describe. Vendors arrive with confident demos; your team can’t answer questions about how the work actually happens, because the answers were never written.
PROBLEM 04
Documentation culture is uneven across departments.
Some teams have wikis going back fifteen years. Others have a binder. Most have neither. Strategy demands a shared baseline before it demands a roadmap.
02 / The method
Qualitative research is the discovery method. The library is the curriculum.
Five phases, in order. The work begins with listening, moves into shared documentation, and reaches tooling only after the workflow has been mapped. Human judgment is preserved at every step; the deliverables outlast the engagement.
PHASE 01
Listen
Semi-structured interviews with staff at every level — circulation, cataloging, reference, IT, leadership. The language people actually use to describe their work becomes part of the evidence.
2–3 weeks · 18–30 conversationsPHASE 02
Map
Workflows, decision points, and hand-offs are diagrammed — not only as documented, but as they actually run. Annotated, dated, attributed.
Workflow atlas · decision logsPHASE 03
Synthesize
Scattered process knowledge becomes a shared documentation layer your team can edit, contest, and extend. The map belongs to the institution.
Knowledge library · review cyclesPHASE 04
Equip
With the work mapped, LibraryOps assesses where AI and automation responsibly fit — and, just as often, where they should not. Every recommendation cites the workflow it touches.
AI readiness reviewPHASE 05
Sustain
The final phase establishes rituals, owners, and review cadences so the operational memory keeps growing after the engagement ends.
Playbooks · quarterly check-ins
03 / Services
Engagements scoped to your operational reality, not a fixed product.
Start with an Operational Memory Audit. It creates a shared picture of the work, the knowledge at risk, and the right next step before a larger engagement is considered.
Illustrative first pilot
One department. One fragile workflow. One shared map. A focused pilot might document local cataloging exceptions, an interlibrary-loan escalation path, or the onboarding knowledge held by one experienced staff member.
START HERE01
Operational memory audit
A focused intake of the institution’s undocumented work. LibraryOps maps where knowledge lives, where it’s thinning, and where a single retirement would create real exposure.
- Stakeholder interviews
- Knowledge-risk register
- Executive readout
SERVICE02
Workflow mapping engagement
Department-by-department workflow atlases drawn from observation and interview. The map your successors wish you had left them.
- Annotated process diagrams
- Decision log capture
- Exception handling traces
SERVICE03
Documentation culture program
The hard part isn’t writing the documentation. It’s the rituals that keep it alive. LibraryOps designs the ownership, cadence, and review practices that hold.
- Owner & cadence design
- Templates & rubrics
- Onboarding integration
SERVICE04
Responsible AI readiness review
Before any model touches your operations, LibraryOps assesses what’s ready, what isn’t, and where automation would damage trust. Every finding cites the workflow it concerns.
- Capability mapping
- Boundary recommendations
- Vendor evaluation rubric
SERVICE05
Leadership advisory retainer
A standing relationship for directors and deans navigating modernization. Quarterly working sessions, on-call strategic review, and an ongoing thinking partner.
- Quarterly working sessions
- Decision-stage reviews
- Confidential sounding board
SERVICE06
Succession & transition support
When senior staff retire, restructure, or move on, a focused capture engagement preserves working knowledge for the people who will carry it forward.
- Departing-staff interviews
- Knowledge transfer plans
- Successor briefing kits
04 / Deliverables
Artifacts your team owns, edits, and keeps growing.
Engagements are designed to produce concrete, editable deliverables — not a slide deck that goes stale. The people who do the work can review, contest, and keep extending them.
D.01
Markdown + SVG
Workflow atlas
Annotated diagrams of the actual process — including the exceptions, escalations, and hand-offs that the formal procedure misses.
D.02
Structured log
Decision log
A dated record of operational decisions and the reasoning behind them. Backfilled from interviews; maintained going forward.
D.03
Wiki-ready
Process knowledge library
A searchable repository of the tacit knowledge that previously lived in inboxes and tenured staff memory.
D.04
Report + matrix
AI readiness assessment
A workflow-by-workflow review of where automation responsibly fits, where it doesn’t, and what would need to change first.
D.05
Operating guide
Documentation playbook
The rituals, templates, and ownership patterns that keep the institutional memory alive after the engagement ends.
D.06
Living register
Knowledge-risk register
A standing inventory of where institutional knowledge is concentrated in too few people, with mitigation plans.
FIG.02 — deliverables map
05 / Responsible AI readiness
AI systems are the implementation layer, not the headline.
LibraryOps works with AI infrastructure without treating it as the answer to every problem. The principles below define the boundary for every recommendation.
P.01
Map before you model.
No automation recommendation without a documented workflow underneath it. The map is the prerequisite, not the deliverable.
P.02
Automate analysis, not care.
Patron experience, judgment calls, and the relational work of librarianship stay human. Vendor proposals are evaluated against that boundary.
P.03
Preserve human judgment at every checkpoint.
Models can summarize, surface, and suggest. They do not decide. Every equipped workflow requires a named human who signs off.
P.04
Document the work first.
If a workflow can’t be described to a junior staff member, it cannot be responsibly automated. Documentation is the gate, not the bonus.
P.05
Vendor claims are evidence, not conclusions.
LibraryOps brings the rubric. Demos are evaluated against your real workflows, not a vendor’s sample data.
P.06
Audit trails are infrastructure.
Every automated step needs to be reviewable, reversible, and attributable. If it isn’t, it isn’t ready.
The line
Automate analysis and support. Do not automate away care, trust, human judgment, or the human-centered patron experience.
06 / Collaboration
LibraryOps works alongside the whole institution, not just the org chart’s top.
Different audiences need different conversations. Pick the one that fits where you sit.
“Before we modernize, we need to know what we actually do.”
Library leadership often carries the strategic question alone. LibraryOps helps answer it with evidence drawn from your own staff and patrons, not from a sector report.
Advisory retainers can follow the cadence of academic, public, and special-collection leadership: quarterly working sessions, decision-stage review, and confidential sounding-board access in between.
What you leave with
- A defensible operational picture
- An AI readiness review you can take to the board
- A documentation practice that survives the next leadership transition
“The vendor wants requirements. We need workflows first.”
IT and systems collaborators are often handed a modernization mandate without the operational map underneath it. LibraryOps closes that gap with workflow documentation that supports specification, evaluation, and integration.
The practice does not replace your team. It contributes the qualitative discovery work that systems implementations rarely budget for and that quietly determines whether the project succeeds.
What you leave with
- Workflow specifications grounded in observed practice
- Vendor evaluation rubrics tied to real exceptions
- A shared vocabulary with library leadership
“Finally, someone is writing down what we actually do.”
The interview method centers people who have been doing the work for years and may never have been asked to describe it. LibraryOps comes to listen. The map belongs to the institution, and the words in it belong to the staff.
The process is non-extractive. Staff review every diagram before it is finalized; nothing is published or shared without explicit consent.
What you leave with
- A documented version of work that’s been invisible
- A successor who won’t start from zero
- A seat at the AI readiness conversation
“Show me how this institution will hold its knowledge through the next decade.”
Trustees, board chairs, and funders increasingly ask whether the libraries they support are positioned for the next ten years. LibraryOps produces the evidence base leadership needs to answer without resorting to vendor talking points.
For grant cycles and strategic-plan updates, the deliverables can stand alone — honest about what works, what does not, and what the institution still needs to decide.
What you leave with
- An evidence-backed strategic posture
- A clear, ethical AI position
- Documentation suitable for grant and accreditation cycles
FAQ / Working together
Clear boundaries before the work begins.
The method depends on trust, staff participation, and institutional ownership. These are the starting assumptions.
Does LibraryOps automate library work?
Not by default. The work begins by documenting the real workflow. Automation is considered only where it supports staff judgment, access, and accountability.
How is staff knowledge handled?
Staff review representations of their work before anything is finalized. Confidentiality, attribution, access, and sharing boundaries are agreed at the start.
What does a first engagement require?
A defined department or workflow, access to relevant staff and existing documentation, a leadership sponsor, and time for review. A focused pilot can stay deliberately small.
07 / Start a conversation
Start with a direct email.
Click the address below to copy it. Then paste it into a new message in whichever email service or app you use.
After copying, paste the address into your email compose window.
PRACTICE DETAILS
Engagement length6–14 weeks
Cohort size3 institutions / quarter
Interview scope18–30 staff conversations
Deliverable formatMarkdown, SVG, structured logs
Designed forAcademic, public, special
AvailabilityRolling conversations