Catalog the whole library, copy by copy. ISBN, DDC, accession, shelf, condition.
Every book is one record. Every copy is its own row with an accession number, barcode and shelf location. Authors are kept clean across the whole library. Categories are hierarchical, in your school's own DDC scheme or a custom subject tree. Schools with 800 books or 80,000 books use the same screens.
How most school libraries get catalogued (and why it stops mattering by year three)
A new librarian walks into a CBSE school in Indore in June. The previous librarian retired in March, took her register with her, and what is left is a wooden cupboard with about 4,200 books, a half-finished Excel sheet with 2,800 of them, and an annual donation drive box from last winter that nobody opened. Class 6 reading period starts on Monday. Three classes have been scheduled into the library and the new librarian has to issue books to children whose parents are paying tuition for a working library.
The Excel has columns for title, author and class. It does not have ISBN. It does not have shelf location. It does not have a record of how many copies of each book the school owns. It does not link the same author across two books — Wings of Fire by A.P.J. Abdul Kalam is one row; Ignited Minds by A.P.J. Abdul Kalam is another, and the spelling is slightly different in each. Replacing this with a real catalog takes the new librarian eight months of weekends, by which point another year of donations has arrived and the gap is wider than when she started.
The gap is not the librarian's fault. The gap is that a flat spreadsheet treats every book as one isolated row when in reality a book has three relationships. It has many copies. It has one or many authors. It belongs to a category in a hierarchy. Every legacy library system that flattens these relationships forces the librarian to maintain duplication by hand. Inkwelly does not.
How the catalog works on Inkwelly
The catalog has three screens that mirror how a librarian thinks: Books, Copies, Authors. Plus a Categories tree.
Books. A book record is one work — the title, subtitle, ISBN-10, ISBN-13, publisher, year, edition, language, format (hardcover, paperback, ebook, audiobook, magazine, journal, reference), pages, description, summary, DDC number, your school's call number, tags, cover image, unit price. The book record is the metadata you would put on a Goodreads page. It does not represent any specific physical copy.
Copies. A copy is one physical specimen — its accession number, barcode, shelf location, condition (new, good, fair, poor, damaged, lost), current state (available, issued, reserved, lost, damaged, withdrawn), purchase date, purchase price, source vendor, free-text remarks. A school with 6 copies of Beehive (NCERT Class 9 English) has 6 copies, each with its own accession number and condition, all hanging off the same book record.
Authors. An author is one entity, normalised across all books they wrote. Many books share the same author — think NCERT books crediting the NCERT committee, Panchatantra and Akbar Birbal both attributed to Indian folklore. Authors are joined to books in order, so a co-authored book preserves Russell, Stuart and Norvig, Peter the right way around. Renaming an author once cleans up every book they appear on — no more A.P.J. Abdul Kalam vs Dr A P J Abdul Kalam vs Kalam, A. P. J. drift.
Most cataloging happens on the books screen, because that is where the librarian thinks. Adding copies is a side panel on the book. Authors is a list. Categories is a tree.
What every book record holds
- Title and subtitle — the visible name on the spine, plus the long subtitle for academic texts
- ISBN-10 and ISBN-13 — both formats stored, so a 1990s book without ISBN-13 still indexes cleanly
- Publisher, year and edition — the citation triad; helps when buying a replacement copy
- Language — free text, so Hindi, English, Sanskrit, Urdu, Tamil, Malayalam etc. all work as the school catalogs them
- Format — hardcover, paperback, ebook, audiobook, magazine, journal, reference, or other
- Page count — used for reading-time estimates and reading-level alignment to NEP 2020 reading lists
- DDC number and call number — Dewey Decimal Classification plus the school's own shelving call number; both optional
- Tags — free-form, useful for theme tags like NEP-Recommended, Annual-Function-Prize, State-Board-Class-9
- Cover image — attach from the media library so the screen looks like a real catalog, not a spreadsheet
- Unit price — used for lost and damaged book replacement charges; supports any currency
What cataloging actually looks like
Bulk accession with prefix and counter — the only realistic way to onboard 5,000 books
Most schools onboarding to Inkwelly already own a few thousand books. Typing accession numbers one by one is not realistic. The bulk accession flow uses three small settings: your school's prefix (e.g., AVM/2026/), the next number, and the zero-pad width (default 6, so the first accession is AVM/2026/000001).
The librarian goes book by book, picks 'Add 12 copies', and Inkwelly creates AVM/2026/000247 through AVM/2026/000258 in a single click. The counter advances on its own. Existing books retain whatever accession numbers they had — no system rewrites your historical numbering. New copies pick up where the counter is. The CSV import flow accepts pre-numbered copies for schools that have already digitised the spreadsheet but want to migrate without touching their accession scheme.
Categories are hierarchical, your scheme, your call
The categories tree is fully recursive — every category can have a parent. Most schools build a small DDC-style tree: 000 General Works, 100 Philosophy, 200 Religion, 300 Social Sciences, 400 Languages, 500 Pure Sciences, 600 Technology, 700 Arts, 800 Literature, 900 History and Geography. Inside each, sub-categories: 510 Mathematics, 530 Physics, 540 Chemistry, 570 Life Sciences. Some primary-section libraries skip DDC entirely and build their own tree: Story Books, Picture Books, Reference, Atlases, Hindi Section, Sanskrit Shlokas.
The tree is drag-and-drop. Books can be assigned to any node. Reports roll up totals to the root — the librarian sees "832 books in 500 Pure Sciences" without manually summing the children. The category structure is fully your school's choice, not Inkwelly's. We do not ship a default DDC tree on first install — most school librarians have a strong opinion about how their library is organised, and we respect that.
Authors normalised, not duplicated
A single author record stores a name, optional biography and optional nationality. Books link to authors in order, so Russell, Stuart and Norvig, Peter on a Class 12 Computer Science textbook displays in that order, every time, in the catalog and on the spine label printout.
When the librarian misspells A.P.J. Abdul Kalam in one book and APJ Abdul Kalam in another, they merge the two author entries from the Authors screen — every book the merged author was on automatically points to the canonical entry. No book record needs editing. The cleanup is one action, not 30. This is the single biggest reason school librarians say their second-year cataloging is faster than their first — the duplication that legacy systems accumulate, Inkwelly never accumulates.
Each copy carries its own state — condition, shelf, vendor, price
The library owns three copies of Hornbill. Copy A is on the shelf, condition good, accession AVM/2026/001284. Copy B is issued to a Class 11 student, condition good, due back in nine days. Copy C is in the damaged-shelf state because last term someone tore page 47, condition damaged, currently waiting for a write-off decision.
Three copies. One book. The librarian sees all three on one screen. The available count on the books list reads "2 of 3 available". The reservation queue knows about the issued copy. The fine engine knows about the damaged one (its replacement charge was applied at return). The shelf-audit report can list every available copy and where it is supposed to be — useful when the librarian does the half-yearly stock take and finds three books that are physically present but listed as lost.
“First time I had a real catalog — 12,000 copies, no duplicate authors, no missing accession numbers. The auditor from CBSE didn't ask any follow-up question. That's the only review that matters.”
Real cataloging situations the system handles
-
A donation drive arrives with 380 books, none of them in the catalog. The librarian opens bulk add, types one row per book (title, ISBN if available, copies count), the system creates the book and copy records, accession numbers are auto-assigned. A printable barcode sheet is generated for the same accession numbers — stick them on the spines over a long weekend.
-
The school owns 3 copies of Hornbill but the catalog says 1. The librarian opens the book record, hits 'Add 2 copies', specifies accession numbers (or auto-generates), and the available count updates from 1 to 3 immediately. Search results always show the right available count without the librarian doing any extra work.
-
Two duplicate book records exist for The Diary of a Young Girl. The librarian opens the merge dialog on the books list, picks the canonical record, and the second record's copies are reparented to the first. The duplicate is hidden from the live screens but its history is preserved. Reservation queues, issue history and fines all stay attached to whichever original record they were on — nothing is lost, nothing is double-counted.
-
The school migrates from a previous library system with 14,500 books in CSV. The CSV import accepts a 17-column format (book metadata + author array + category code + copies array). Inkwelly validates ISBNs, dedupes authors against existing entries, creates the missing categories, and produces an import report with rows that succeeded, rows that failed and rows that need manual review. A 14,500-book import typically takes 9 to 12 minutes.
-
The librarian wants a printable label for every copy on a specific shelf. The shelf-location filter on the copies screen narrows to all copies in shelf A1, the print barcode sheet flow generates an A4 PDF with one barcode per copy, formatted at 30 labels per page (3 columns by 10 rows). The label sticker pack is the standard 3M / Avery 5160 size, available at any stationer.
Operations the catalog supports out of the box
- Add a book in 12 keystrokes (title + class category) and enrich later
- Bulk add multiple copies of the same book — one click, sequential accession numbers
- Hierarchical categories with drag-and-drop reorder and parent reassignment
- Author merge — collapse misspellings into one canonical entry, all books re-link
- ISBN search across the catalog — ISBN-10 and ISBN-13 both match, with hyphenated variants
- Cover image from the media library or upload directly during catalog entry
- Printable barcodes — one PDF per shelf, one PDF per accession range or one PDF for new arrivals
- Soft delete — nothing is hard-deleted, history is preserved forever
- CSV export of the entire catalog for external reporting or affiliation submissions
Catalog your library on real data, in your own session
30-minute walkthrough using your existing book Excel or paper register. We will load 100 books on screen with you watching, then hand you the keys.
Limits, scale and the small print
Soft-delete by default. Books, copies, authors and categories all have a delete flag. A removed entry is not gone — it is invisible to the live screens but visible to the audit history. The deleted entry's issues, reservations and renewals are preserved. Nothing is hard-deleted by the librarian; only an org-admin can hard-delete with a separate confirmation flow.
Counts are kept fresh automatically. The book row's total copies and available copies stay in sync as books are issued, returned, lost or damaged. They are also re-checked daily by a reconciliation routine. If you ever see a drift, run the reconciliation manually — the report tells you which book row was off and by how much.
Scale. The catalog has been tested with 240,000 books and 1.1 million copies in a single library, with sub-200ms search response times. ISBN, title, author, category, DDC number and accession number are all indexed. There is no expected scale ceiling for any school in India — even the largest residential school libraries we have onboarded sit well below the limits.
Multi-library. A school can run several libraries — a primary library and a senior library, a sports library and an academic library, an English-medium library and a Hindi-medium library. Each library has its own catalog. Books cannot move between libraries automatically — a transfer is an explicit action with a transfer reason, recorded in the transfer audit log. This matches how schools actually run their internal libraries and avoids the silent moves that legacy ERPs allow.
Data residency and export. All catalog data is stored on Mumbai-region servers, complying with the DPDP Act 2023. The full catalog is exportable as CSV any time. Cover images live on Cloudflare R2 with content-hashed URLs so old image links always continue to work; new uploads get fresh URLs.
Belongs to
1 moduleFrequently asked
7 questionsDo we have to use Dewey Decimal Classification, or can we build our own subject tree?
Either. The categories tree is fully hierarchical, so schools can build a DDC tree, a custom subject tree, or even a hybrid (DDC for senior section, custom for primary). Inkwelly does not ship a default tree — we let the school build what makes sense for their library.
Can we import an existing CSV or Excel of our books?
Yes. The CSV import accepts a 17-column format covering book metadata, author array, category code and copy array. The import dedupes authors against existing entries, validates ISBNs and produces a report with succeeded, failed and needs-review rows. A 14,500-book import typically takes 9 to 12 minutes.
Are accession numbers auto-generated, or can we keep our existing numbering?
Both. New copies pick up auto-generated numbers from your prefix and counter — e.g., AVM/2026/000247. Existing books migrated via CSV keep whatever accession numbers they had. Schools moving across never have to renumber their historical collection.
How does Inkwelly handle the same book having three physical copies?
Each physical copy is a separate entry hanging off the same book record. Each copy has its own accession number, barcode, shelf location, condition and state. The book row tracks the total and available counts that update automatically as copies are issued, returned, lost or damaged.
Can two books share the same author and have his name updated centrally?
Yes. Authors are kept clean across the whole library and joined to books in order. Renaming an author or merging two duplicates updates every book they appear on automatically — no manual edit on the book records.
Does the catalog support cover images, and where are they stored?
Yes. Each book record has a cover image attached from the school's [media library](/modules/media). Cover images are stored on a CDN with content-hashed paths, so old URLs never break and new uploads get fresh URLs. The catalog screens render covers in card view; the issue desk uses them as a quick visual confirmation when scanning.
Can a school run two separate libraries — say a primary library and a senior library — with completely different catalogs?
Yes. Each library has its own catalog, settings, member categories and circulation log. Books and copies belong to one library; transferring a book between libraries is an explicit action recorded in the transfer audit log so the principal can always see where every copy has lived during its life.
You might also like
2 readsSee Inkwelly on your school
30-minute demo. We open your current ERP with you and load your data into Inkwelly on the call. Dated go-live plan by the end of it.