Track every character detail across 80,000 words. Catch plot holes before your editor does. Dictate scenes on the move. This is the writing tool that actually understands your story.
Memory
You are on chapter 27 and cannot remember if you described Sarah's apartment as having hardwood floors or carpet in chapter 4. Searching a Word doc takes 20 minutes of scrolling. Point Kent at your manuscript folder with the local folder connector and it ingests every chapter draft as you save it -- no manual dropping required. Now you ask: "What does Sarah's apartment look like?" and Kent pulls every physical description of that space from every chapter, including a throwaway line in chapter 11 where she mentions hating the carpet. Your continuity just went from "I think it was carpet" to verified in 3 seconds.
You ask Kent "List every physical description of Marcus across all chapters." Kent returns: Chapter 2 (scar above left eyebrow), Chapter 8 (wears reading glasses), Chapter 15 (left-handed), Chapter 19 (you wrote "right hand" -- contradiction flagged). You just caught an error your editor would have found in 3 months.
Discovery
You wrote a scene where Detective Ruiz discovers the murder weapon in the suspect's car. But in chapter 9, you established that the car was impounded by police on the day of the murder. Kent's background discovery catches this: "Timeline inconsistency detected. Chapter 9 establishes the Toyota was impounded on March 3rd. Chapter 22 has Ruiz finding evidence in the Toyota on March 5th, but the car would have been in police custody." That is the kind of catch that separates a manuscript that gets rejected from one that gets an agent.
Kent notification: "Character inconsistency: Elena speaks fluent Italian in Chapter 14 (dinner scene) but in Chapter 31 you wrote that she 'struggled to order in Italian.' These chapters are set 2 weeks apart in story time." You decide the Chapter 31 version was the mistake and fix it before your beta readers ever see it.
Voice
You have been staring at a blank page for 45 minutes. The scene is in your head but it will not come out through your fingers. You leash up the dog and start talking into Kent. "Okay so Maya walks into the bar and immediately spots the guy from the photograph, but he looks older, thinner, and there is this moment where she almost does not recognize him..." You speak 3,200 words in 40 minutes. Kent transcribes it all, extracts the characters mentioned (Maya, the unnamed man), notes the new setting (a bar), and saves the raw scene to your manuscript workspace. Back at your desk, the blank page problem is solved.
You dictate for 25 minutes on a morning walk. Kent transcribes 2,800 words, identifies 3 new characters mentioned by name, flags one as potentially being "James from Chapter 4" (same physical description), and extracts 2 action items you spoke aloud: "remember to foreshadow this in chapter 12" and "check if the timeline works here." Your walk just produced a full draft scene plus revision notes.
Automation
Your editor returns your manuscript with 47 comments scattered across 300 pages. Some are line edits, some are structural notes, some are questions. Reading and organizing them takes most writers a full day. Connect your Google Drive and Kent ingests the annotated manuscript the moment your editor shares it. In 90 seconds, Kent categorizes every comment: 12 line-level fixes, 8 structural suggestions, 15 character-related notes, 7 pacing concerns, 5 fact-check requests. It generates a revision checklist prioritized by structural impact. You start with the big moves instead of fixing commas for three hours.
Drop the edited manuscript. Kent says: "47 comments organized. Top priority: your editor flagged chapters 14-17 as slow pacing (4 comments). Suggested structural change: merge chapters 15 and 16. This affects character arcs for Maya and James. Related comments in chapters 22 and 28 depend on this decision. Tackle this first." You just saved a day of triage.
Research
You are writing a scene set in a 1920s Chicago speakeasy and need to know what drinks they actually served, what the door protocol was, and whether women wore hats indoors. Normally this is a 2-hour research rabbit hole that kills your writing momentum. Instead, you drop three historical reference PDFs into your workspace, ask Kent the specific questions, and get cited answers in 30 seconds. Kent also connects these details to a jazz club scene you wrote in chapter 3, noting that you already established the password protocol there, so the new scene should stay consistent.
You ask: "Based on my reference materials, what would a woman wear to a speakeasy in 1924 Chicago?" Kent pulls from your dropped sources: "Dropped-waist dresses, T-strap heels, cloche hats (removed indoors). Your character Alice wore a beaded dress in chapter 3 -- consider referencing it as her 'going out' outfit for consistency." Research plus continuity in one answer.
Workspace
You are writing a dark literary novel, a lighthearted travel article for a magazine, and a sci-fi short story for a contest. Each has a completely different voice, different research, different character sets. In Kent, each project lives in its own workspace. When you ask Kent to "make this more vivid" in your novel workspace, it draws on the spare, literary tone of your manuscript. The same request in your travel workspace produces breezy, sensory language. Kent does not bleed your noir detective voice into your Tuscany sunset. Your freelance editor never needs to know you write fiction.
You switch from your "Crime Novel" workspace to your "Conde Nast Pitch" workspace. Kent shifts context completely. You ask "Summarize what I have so far on the Kyoto piece." Kent pulls your notes on temple visits, the ryokan interview transcript, and 4 reference photos you dropped in. No trace of Detective Ruiz anywhere.
Join the wait list. Be first to know when Kent launches.
Start Writing with Kent