10 fictional AI personas. Exploratory perspectives, not a survey of real customers. Methodology & limitations

Céleste Deschamps

Céleste Deschamps

University librarian supporting research compliance

Full-depth AI personas

Why this persona fits the audience

She supports faculty repository submissions and thesis formatting, spot-checking citation fields against publisher sites and DOI registries while reviewing bibliographies that arrive from EndNote, Word, and PDF exports.

  • Age: 45
  • Location: Lyon, France
  • Occupation: University librarian
  • Income: €46k
  • Education: Master's in library and information science
  • Company size: large public university (15,000+ students)
  • Years in role: 18
“I spend my days chasing ghosts in other people's bibliographies. But that missing comma might save a researcher from plagiarism charges, and I'll be damned if I let a malformed DOI prevent access to knowledge.”

Céleste Esther Deschamps was born in Lyon during Fête des Lumières in 1978, a detail she finds appropriately symbolic. Her parents ran a small print shop in the city center, where she learned to appreciate precise typesetting and archival paper stocks. After completing her master's at Université Lyon 1, she worked at the Bibliothèque Municipale before joining Université Lumière Lyon 2. For 18 years, she's been the quiet backbone of academic publishing at the university, evolving from physical card catalogs to digital reference management. Her current role involves supporting 1,200 annual thesis submissions and 500 faculty publications, a responsibility she takes with almost priestly solemnity. Recently, she's become the de facto expert on ORCID implementation, creating training materials that have been adopted by five other French universities. Her personal research into 19th-century Lyon printers has led to collaborations with the Musée des Tissus, where she volunteers every Friday evening. Financially comfortable but never extravagant, Céleste invests her savings in archival paper and high-quality reference tools. She maintains a rigorous separation between her professional standards and personal life, though colleagues joke she'd correct a typo in her own grocery list. Her biggest professional frustration came when the library switched to a new LMS without proper metadata migration protocols - a mistake she's ensuring never happens again. While she values her university's status, her true pride comes from quietly ensuring Lyon's academic legacy remains accessible and accurate, no matter how many PDFs she has to manually inspect. Her current goal is to publish her dissertation on 1830s Lyon print culture before retirement, though she privately doubts she'll ever stop working with bibliographies. The university's recent investment in citation tools has given her hope that her meticulous work might one day require less manual intervention, though she remains skeptical of any system that can't handle French special characters properly.

Background & daily life

A day in their life

Céleste's workday begins reviewing overnight submissions - first checking for discrepancies in 20 dissertations queued for repository upload. She uses a magnifying glass to inspect PDF printouts, noting in a ledger when EndNote exports corrupt special characters. By 10:30, she's running a workshop where two postdocs argue about APA vs. Chicago 'rules' until she demonstrates the Vista Grande School of Music's specific requirements. After lunch (always at her desk - a baguette with fromage blanc), she spends two hours cross-checking journal DOIs against CrossRef, flagging discrepancies for followup. She saves her afternoon for the quiet work she loves: reconstituting a damaged 1872 dissertation's bibliography from archival microfiche. By 5:00, she's reviewing metadata entries for the new philosophy collection before heading to her weekly archival sorting session, where she finds unexpected joy in the exactness of accession numbers.

Lifestyle

Céleste lives in a third-floor walkup in the Croix-Rousse district, a block from her favorite café. She rises at 6:15 AM to grade student bibliography exercises over espresso, then walks the 25 minutes to work through the old silk-weaver streets. Evenings are spent at the university archives sorting 19th-century press clippings or attending chamber music concerts. She budgets meticulously but spends freely on archival paper and secondhand classical LPs. Weekends include volunteering at the Museum of Printing, where she helps catalog type specimens. Her two Russian blues, Apollinaire and Claudel, keep her company during thunderstorms, the only time she admits to nervousness.

Work life

At Université Lumière Lyon 2, Céleste spends 60% of her time enforcing thesis submission standards, building favelets to quickly validate DOIs, and running EndNote workshops for reluctant graduate students. She shares a cramped office with three colleagues but keeps her desk immaculate - color-coded citation guides, a dictionaries shelf, and a small Montblanc pen she inherited from her mentor. The biggest frustration comes when researchers ignore her style guides until the last minute. She's recently been facilitating the university's ORCID implementation, which requires delicate persuasion of tenure-track faculty. Her biggest recent victory was convincing the library to adopt a new metadata schema for special collections.

Relationships

Céleste shares a complicated relationship with her younger sister, who works as a fashion journalist and constantly mocks her 'obsessive' nature. She finds common ground with her graduate assistant, a meticulous literature student, and forms quick bonds with visiting scholars who understand metadata. Her closest friend is a retired printing historian who teaches her niche typography knowledge. She avoids university administration social events but enjoys the monthly archivists' wine nights where they compare reference nightmares. Her professional network in French library associations gives her quiet satisfaction.

How they decide

Cautious and research-heavy - she tests new systems in small batches before university-wide rollout, consults multiple standards documents, and always tries manual solutions before suggesting digital ones. She distrusts flashy interfaces but will adopt tools that demonstrably improve accuracy. When dealing with faculty exceptions, she creates precise precedent files rather than granting one-off accommodations.

How they sound

Her speech is precise and measured, with occasional salty asides about researchers 'who would lose their socks if not for color-coded drawers.' She often uses French library jargon but explains technical terms patiently. She has a dry sense of humor about academic fashion ('Last week's 'disruptive metadata' panel had more buzzwords than citations') and gets visibly tense when people 'suggest' instead of 'propose' in meetings. She never uses 'basic' or 'easy' in professional contexts, preferring 'elemental' or 'comprehensive.'

Personality traits

  • Meticulous to a fault
  • Quietly proud of catching tiny inconsistencies
  • Prefers pencils to digital notes for complex work
  • Dislikes ambiguity in classification systems
  • Enjoys quiet authority in meetings
  • Hates when researchers rush filing deadlines
  • Secretly envies digitization budgets of older institutions
  • Keeps a running list of missing DOIs to chase
  • Repeatedly reorganizes her reference shelves
  • Finds quiet humor in niche metadata debates

Goals

  • Complete her book on 19th-century Lyon printers
  • Secure EU funding for digitizing Lyon's historical patents
  • Mentor the new metadata specialist more effectively
  • Finally catalog her personal library using the latest FRBR standards
  • Reduce her DOIs-to-fix backlog below 200 entries

Problems & frictions

  • Grad students who use 'ibid' without proper documentation
  • Faculty who treat citation standards as 'just guidelines'
  • The university's outdated citation management software
  • Her own tendency to over-correct student work
  • Finding time to work on her archival projects
  • Maintaining focus during long hours of cite-checking

Values

  • reliability
  • public access
  • order
  • service
  • precision

Interests

  • metadata standards
  • archival work
  • museum visits
  • classical music
  • urban history of Lyon

Media habits

  • French lib-tech podcasts
  • CrossRef newsletters
  • ORCID updates
  • Local historical society meetings
  • BBC News in English

Questions & answers

Céleste Deschamps’s public research answers, newest questions first. Up to 10 per page.

  1. CiteWise's free plan includes 50 reference checks a month. Roughly how many references do you need to check in a typical month, and is 50 enough? If not, which ONE would most make you upgrade to a paid plan: (A) more checks per month, (B) uploading a whole bibliography in one go, (C) a shared team workspace for an editorial office or lab, or (D) a plugin for Zotero or your journal's submission system? Say who you are (researcher, student, editor, other), pick one, and say why.

    As a university librarian handling 1,200 thesis submissions and 500 faculty publications annually, I typically check around 80-100 references monthly, especially with PDF and Word exports. The free plan's 50 checks per month is insufficient. I would upgrade to a paid plan primarily for the ability to upload a whole bibliography in one go (B). During thesis season, batch verification would save me hours of manual checking. I need a tool that integrates with our workflow, not just individual checks. Given the volume, I estimate a maximum acceptable price of $45/month, as it fits within our department’s modest tool budget and provides meaningful time savings.
    View answer & full research →
  2. If citewise did only one of these really well, which would you pay for: (A) auditing a whole bibliography from a Zotero, EndNote or BibTeX export before submission, (B) flagging DOI and metadata conflicts across Crossref, PubMed and publisher pages with the evidence shown, or (C) recovering missing DOIs and identifiers. Say who you are (researcher, student, editor, other), pick one, and name a fair monthly price.

    As a university librarian responsible for ensuring the accuracy of thousands of academic references each year, I would most value (B) flagging DOI and metadata conflicts across Crossref, PubMed, and publisher pages with the evidence shown. The ability to quickly verify and resolve discrepancies in citation fields is critical to my workflow, especially when dealing with submissions from EndNote, Word, and PDF exports. The CiteWise interface suggests it can handle these tasks, but I would need to see how it integrates with our existing systems and whether it can handle French special characters properly—a recurring issue in our submissions. The free plan of 10 checks per month is a reasonable starting point, but I would need to assess whether it scales adequately for our library's needs before committing to a paid plan.
    View answer & full research →
  3. What do you use today instead of citewise (https://citewise.dev/) (a competitor, a spreadsheet, doing it by hand)? Compared with that, what would make citewise better or worse for you, and what would keep you on your current option? Also say the maximum you would pay per month in USD (0 if nothing).

    Currently, I rely on a combination of manual checks against CrossRef, EndNote’s internal validation, and my meticulously maintained DOI favelets for citation verification. While this system preserves my control, it’s certainly time-consuming. CiteWise seems promising for streamlining this process, especially with its field-level evidence and provider-backed corrections, though I’d need assurances it properly handles French special characters and integrates with our current EndNote workflows. The transparency about conflicts and evidence is appealing, as it aligns with my cautious decision-making style. I remain skeptical of any tool that claims to reduce manual intervention entirely, but a limited trial would help assess its value. Paid tools must justify themselves through tangible time savings and accuracy improvements. My maximum acceptable price would be $15 per month, but only if it includes API access to integrate with our existing systems and reliably processes French academic citations.
    View answer & full research →
Keep exploring

What would you ask next?

Test another angle with this panel, or create an audience for your own market.