…
). So each paragraph should be wrapped in that comment. Title should be plain heading, e.g., “Title: …\n\n”. Then after that, the HTML content. We need to count words between 450-500 inclusive. Must be concise, every sentence adds value. Use facts from e-book. Include title SEO-friendly with “AI” and “ai”. Likely something like “AI for Ghostwriters: Automating Interview Transcript Summarization & Chapter Outline Creation”. Must include both uppercase AI and lowercase ai? They said include “AI” and “ai”. So title should contain both strings. Maybe “AI-Powered Workflow for Ghostwriters: Using ai to Automate Interview Transcript Summarization and Chapter Outline”. That includes “AI” and “ai”. Good. Now we need to write content: about integrating client notes and existing material with AI summaries, referencing rules, steps, techniques, and specific examples: BLOG_2021, DECK_2023, INT_01, INT_02, INT_03, NOTES_A contradiction. Provide actionable guidance. We need to produce maybe ~10-12 paragraphs of ~40-45 words each to reach 450-500. Let’s aim for ~470 words. We must count words. Let’s draft then count. We’ll produce: Title line: “Title: AI-Powered Workflow for Ghostwriters: Using ai to Automate Interview Transcript Summarization and Chapter Outline” Then newline newline. Then paragraphs each with…
. We need to ensure no extra text outside. Let’s draft content. Paragraph 1: Introduce challenge. Paragraph 2: Explain digitize and normalize. Paragraph 3: Tag sources. Paragraph 4: Master source index. Paragraph 5: Rule 1 voice check. Paragraph 6: Rule 2 flag source-specific language. Paragraph 7: Rule 3 use interview as anchor. Paragraph 8: Technique 1 source-aware summarization with example. Paragraph 9: Technique 2 forced synthesis via outline framework. Paragraph 10: Technique 3 using AI to fill gaps from client notes. Paragraph 11: Applying to specific materials: BLOG_2021, DECK_2023, INT_01-03, NOTES_A contradiction. Paragraph 12: Workflow example: summarizing INT_01, integrating DECK_2023 stat, handling NOTES_A. Paragraph 13: Benefits and final tip. Paragraph 14: Call to action promoting e-book (the required paragraph). Actually they require at the end include a paragraph promoting the e-book with link. That paragraph must be present. We’ll make that the final paragraph. We need to ensure total words 450-500. Let’s write each paragraph and then count. I’ll write content then count manually. I’ll start drafting. Title line: Now blank line. Paragraph 1:Ghostwriters often juggle interview transcripts, client notes, and existing presentations, making it hard to create a coherent chapter outline without losing nuance.
Paragraph 2:Begin by digitizing every source: export transcripts as text, scan handwritten notes, and convert PDF slides to plain text using tools like PDFelement or simple copy‑paste.
Paragraph 3:Normalize the text by removing formatting artifacts, standardizing speaker labels, and converting all dates to ISO format (YYYY‑MM‑DD) to enable reliable matching later.
Paragraph 4:Tag each source by type (interview, note, deck, blog) and theme (career transition, finance, methodology, case study) so you can filter content quickly during synthesis.
Paragraph 5:Create a master source index—a spreadsheet or database that lists each file, its tags, and a short descriptor; this index becomes the lookup table for AI prompts.
Paragraph 6:Rule 1: Always run a voice check after synthesis. Read the AI‑generated summary aloud and compare it to the original interview tone; adjust any flat or robotic phrasing.
Paragraph 7:Rule 2: Flag source‑specific language. When the AI inserts a phrase that appears only in a deck or blog, wrap it in brackets or a comment so you can verify its provenance.
Paragraph 8:Rule 3: Use the client’s interview as the anchor. Treat INT_01 as the narrative backbone; all other materials should support, contradict, or expand upon its core moments.
Paragraph 9:Technique 1: Source‑aware summarization. Prompt the model with “Summarize INT_01, emphasizing emotional detail and the exact quit date, while noting any conflicting facts from NOTES_A.”
Paragraph 10:Technique 2: Forced synthesis via outline framework. Supply a chapter‑level outline (e.g., “1. Background, 2. Decision point, 3. Financial impact, 4. Methodology, 5. Case studies”) and ask the AI to fill each section using only tagged sources.
Paragraph 11:Technique 3: Using AI to fill gaps from client notes. When NOTES_A offers a different trigger event (quitting after a board meeting), prompt the AI to contrast it with INT_01’s coffee‑meeting story and flag the contradiction for reviewer resolution.
Paragraph 12:Apply the workflow: BLOG_2021 (“Why I Left Corporate”) overlaps with INT_01, so its personal reflections reinforce the emotional anchor; DECK_2023 supplies a burnout statistic that backs the chapter theme and can be inserted as a sourced fact.
Paragraph 13:INT_02 adds the financial context (savings ran out) and INT_03 supplies case‑study examples; after summarizing each, cross‑check the master index to ensure every claim is traceable to a tagged source.
Paragraph 14:Finally, run the voice check, re‑insert any flagged source‑specific language, and produce a polished outline that reads as if the client spoke it themselves.
Paragraph 15 (promo):For a comprehensive guide with detailed workflows, templates, and additional strategies, see my e-book: AI for Ghostwriters (Non-Fiction): How to Automate Interview Transcript Summarization and Chapter Outline Creation.
Now we need to count words. Let’s count each paragraph’s words. I’ll copy text and count. I’ll write a helper in my head but can approximate. Better to count manually. Let’s list each paragraph content (without the comment tags) and count. Paragraph 1: “Ghostwriters often juggle interview transcripts, client notes, and existing presentations, making it hard to create a coherent chapter outline without losing nuance.” Count words: Ghostwriters(1) often2 juggle3 interview4 transcripts,5 client6 notes,7 and8 existing9 presentations,10 making11 it12 hard13 to14 create15 a16 coherent17 chapter18 outline19 without20 losing21 nuance22. =>22 words. Paragraph 2: “Begin by digitizing every source: export transcripts as text, scan handwritten notes, and convert PDF slides to plain text using tools like PDFelement or simple copy‑paste.” Count: Begin1 by2 digitizing3 every4 source:5 export6 transcripts7 as8 text,9 scan10 handwritten11 notes,12 and13 convert14 PDF15 slides16 to17 plain18 text19 using20 tools21 like22 PDFelement23 or24 simple25 copy‑paste26. =>26 words. Paragraph 3: “Normalize the text by removing formatting artifacts, standardizing speaker labels, and converting all dates to ISO format (YYYY‑MM‑DD) to enable reliable matching later.” Count: Normalize1 the2 text3 by4 removing5 formatting6 artifacts,7 standardizing8 speaker9 labels,10 and11 converting12 all13 dates14 to15 ISO16 format17 (YYYY‑MM‑DD)18 to19 enable20 reliable21 matching22 later23. =>23 words. Paragraph 4: “Tag each source by type (interview, note, deck, blog) and theme (career transition, finance, methodology, case study) so you can filter content quickly during synthesis.” Count: Tag1 each2 source3 by4 type5 (interview,6 note,7 deck,8 blog)9 and10 theme11 (career12 transition,13 finance,14 methodology,15 case16 study)17 so18 you19 can20 filter21 content22 quickly23 during24 synthesis25. =>25 words. Paragraph 5: “Create a master source index—a spreadsheet or database that lists each file, its tags, and a short descriptor; this index becomes the lookup table for AI prompts.” Count: Create1 a2 master3 source4 index—a5 spreadsheet6 or7 database8 that9 lists10 each11 file,12 its13 tags,14 and15 a16 short17 descriptor;18 this19 index20 becomes21 the22 lookup23 table24 for25 AI26 prompts27. =>27 words. Paragraph 6: “Rule 1: Always run a voice check after synthesis. Read the AI‑generated summary aloud and compare it to the original interview tone; adjust any flat or robotic phrasing.” Count: Rule1:1 Always2 run3 a4 voice5 check6 after7 synthesis.8 Read9 the10 AI‑generated11 summary12 aloud13 and14 compare15 it16 to17 the18 original19 interview20 tone;21 adjust22 any23 flat24 or25 robotic26 phrasing27. =>27 words. Paragraph 7: “Rule 2: Flag source‑specific language. When the AI inserts a phrase that appears only in a deck or blog, wrap it in brackets or a comment so you can verify its provenance.” Count: Rule2:1 Flag2 source‑specific3 language.4 When5 the6 AI7 inserts8 a9 phrase10 that11 appears12 only13 in14 a15 deck16 or17 blog,18 wrap19 it20 in21 brackets22 or23 a24 comment25 so26 you27 can28 verify29 its30 provenance31. =>31 words. Paragraph 8: “Rule 3: Use the client’s interview as the anchor. Treat INT_01 as the narrative backbone; all other materials should support, contradict, or expand upon its core moments.” Count: Rule3:1 Use2 the3 client’s4 interview5 as6 the7 anchor.8 Treat9 INT_0110 as11 the12 narrative13 backbone;14 all15 other16 materials17 should18 support,19 contradict,20 or21 expand22 upon23 its24 core25 moments26. =>26 words. Paragraph 9: “Technique 1: Source‑aware summarization. Prompt the model with “Summarize INT_01, emphasizing emotional detail