Karen Rodriguez uses ChatGPT Work to turn her Substack subscribers' survey responses into personalized and yearlong reading plans for them to explore classic works of literature.
Karen grew up across the street from a small library in Los Angeles, learning English from Highlights magazines. The classics only became part of her life while she was homeschooling her six children, now ages 10 to 19. Reading Kipling and Twain aloud introduced her to books she hadn't learned to love in school.
After suffering a stroke during her sixth pregnancy, Karen turned to reading to help rebuild her focus, opening books at home while her children slept. Jane Eyre gave her what she calls bibliotherapy: a cathartic way to explore parts of herself through a book. Reading also let her more easily join conversations with people she admired.
Her Substack, Literary Fancy , has grown to about 8,000 subscribers in less than a year, she says. Many want to recover the attention they had lost to scrolling their screens, or to become the kind of person who finishes classics. They also want to join the discussions they see unfolding under her articles (a single post can draw more than 500 comments). To give each reader a starting point, Karen asks what works they have loved or abandoned, what they hope to read, and how much time they have each day. ChatGPT Work retrieves their Google Forms responses from Gmail and uses an agent Karen developed to turn those answers into a custom guide across books she knows readers love. For readers nervous about Russian literature, she starts with Turgenev or Chekhov's short stories before the longer novels.
The plans stretch across 12 months and can run up to 20 pages. They start with shorter classics and scaffold toward harder books. Each recommendation explains what the reader might enjoy about the book, based on their existing interests, and suggests what to look for while reading. Karen refined the agent over several attempts, teaching it to explain its choices through each reader's interests. The guides explain what to look for, suggest swaps for books already read, and offer ways to go deeper. Some readers tell her they cried when they first read their plans, because they finally felt seen.
An MCP integration connects ChatGPT Work to her Substack data, so Karen can ask which posts performed best and develop related posts. Her article series on "unhinged" classics has helped draw readers to her newsletter. These tools let her build a reading business alongside a full-time job and raising six kids.
She encourages readers to put their phones in another room and read for 20 minutes, then write one sentence about what stayed with them. She also suggests stopping at a cliffhanger, even before the chapter ends, so they leave themselves with a question they need to get answered later.
Readers write back each month to tell her about books they’ve finished. When her book club read Wilkie Collins’s The Woman in White, some folks told her they would never have picked it up themselves – but they couldn’t put it down.
