«The Thin Volume» — a book media channel on Telegram focusing on short Russian and translated fiction from the 20th–21st centuries. The focus is on novellas, short stories, small novels, and collections of tales, each under 200 pages. Primary Audience — adult Russian-speaking readers who are returning to reading after a break or seeking a way to fit it in among work and studies. A secondary audience includes people interested in discovering a new author through a short piece and finding others with whom to discuss it. Media Goal — to make the process of choosing a book clearer and help readers gradually return to consistent reading habits. Recommendations consider not only the length of the work but also its mood, narrative pacing, linguistic complexity, and emotional intensity.

Official catalogs from publishers such as Azbuka, AST, Eksmo, and MIF provide essential publication details: author name, title, translator, publisher, year of release, ISBN, page count, and book composition. This data is used to populate the work’s card and verify that the page count adheres to the 200-page limit. Annotations offer a preliminary introduction to the book but do not replace the editor’s reading of the work.

The library catalogs of the Russian State Library and the National Electronic Library are utilized to refine bibliographic information. They assist in locating different editions of a single work, verifying its scope, and supplementing data missing from the publisher’s website. If discrepancies arise, the editor cross-references the records to determine if they belong to the same edition.

The work itself serves as the primary source material for the substantive content of the publications. The editor reads the physical book or a legally acquired electronic version, noting features of the language, composition, and characterization. Based on this, the editor develops original recommendations, topics for supplementary materials, and discussion questions.

The literary media «Polka» and «Prochteniye» are used to familiarize oneself with the historical context, critical interpretations, and interviews. «Polka» primarily aids in working with Russian literature, while «Prochteniye» focuses on reviews and materials about diverse authors. If the publication incorporates another interpretation, the author and source of that interpretation are cited.

The «Thin Volume» project utilizes two types of data processing—machine-driven and human-driven. Combining these methods allows for the consistent release of content while maintaining the author’s unique approach to recommendations.

Machine Processing involves systematizing book information, sorting by theme and volume, eliminating duplicate entries, generating AI drafts, scheduling approved posts, and processing analytics.

Human Processing includes reading and selecting works, verifying sources and publication data, assessing the complexity and nature of the text, formulating recommendations and discussion prompts, and editing and approving publications.

The Division Principle: the machine handles data structuring and repetitive tasks, while the human is responsible for veracity, literary critique, authorial perspective, and alignment of the material with the media concept.

To manage the media workflow, we plan to utilize Google Sheets as the editorial database, n8n to connect processing stages, a language model for structuring materials and drafting content, and a Telegram bot for publishing approved texts.

The service receives curated materials and source links from the editor. Its task is to extract verified information and complete the publication record. For every piece of information, the original source must be retained.The author, title, publisher, year of publication, ISBN, translator, page count, and book composition are recorded in separate fields. The year of the work’s initial publication is specified separately, if confirmed. If information is missing, the service flags it as «No data.» If sources contradict one another, the service saves both options and generates a query for the editor.

The publication service receives the approved text, version number, category, and release date. Before submission, it checks the material’s status, verifies that the current version matches the approved version, and ensures there is no previously published entry with the same identifier.

The system receives metrics entered by the editor for the same period following publication, survey results, and anonymized comments. Quantitative calculations are performed in a table, while the language model groups textual feedback. The summary highlights themes such as book selection, text complexity, publication accessibility, format convenience, and topics for future materials. Each conclusion must be accompanied by the number of responses upon which it is based. Observations must be separated from assumptions.