Edit with
natural language
EditAssist
Connected
Tell EditAssist what to do...
- Find every close-up of the lead actor across all three camera angles.
- Pull all B-roll of the city skyline and arrange by time of day.
- Find the best reaction shot of the interviewer and cut it in after each answer.
- Find the three best takes of the monologue and lay them side by side for the director.
- Search for every clip shot on the Alexa Mini and filter by ISO above 1600.
- Find every exterior shot and check if any are overexposed by more than a stop.
- Find every shot of the red car and create a selects timeline.
- Pull all the behind-the-scenes footage and build a 3-minute BTS reel.
- Find every clip longer than 60 seconds where nobody speaks.
- Find all clips shot at golden hour and arrange them chronologically.
- Build a rough cut from today's interview footage, colour-match to episode 3, and sync the external audio.
- Create a 90-second highlight reel from yesterday's shoot, set to the temp score.
- Build selects from scene 14, prioritising takes the director marked.
- Trim every clip on V1 to remove the first and last two seconds of each take.
- Compare the current cut to yesterday's version and show me what changed.
- Detect every jump cut in the interview and smooth them with cutaways from the B-roll.
- Build a sizzle reel from the best moments across all six episodes.
- Build an assembly from the shooting script: match each scene description to the best take.
- Build a timeline from the storyboard: one shot per panel, with placeholder durations.
- Add the voice-over recording to V2 and time it to the existing picture edit.
- Match the colour grade from the reference still across the entire timeline.
- Apply the CDL values from the colourist's report to every graded clip.
- Grade the night exteriors to match the reference LUT, then render proxies for editorial.
- Apply the show LUT to all clips in the Resolve timeline and render a master ProRes 4444.
- Convert the log footage from the Sony FX6 to Rec.709 using the official S-Log3 LUT.
- Apply a film grain overlay to the entire timeline at 35mm strength.
- Create a side-by-side comparison of the graded and ungraded timeline for client approval.
- Match the colour temperature between the interior A-cam and B-cam setups.
- Open the Resolve timeline, add a power window to the sky, and push it two stops warmer.
- Generate a LUT from the hero shot grade and apply it across the entire reel.
- Sync the boom mic and lavs to the A-cam, then flag any clips with audio drift.
- Replace the temp music with the licensed track and adjust all edit points to the new beat.
- Run a loudness check on the mix and flag anything outside EBU R128 spec.
- Identify every speaker in the podcast and label the timeline markers with their names.
- Detect all the clap sync points in the multi-camera shoot and align the angles.
- Normalize the audio across all interview clips to -23 LUFS for broadcast delivery.
- Match the audio levels between the lapel mic and the boom across the entire scene.
- Detect silence longer than 2 seconds in the podcast and remove it with crossfades.
- Measure the dynamic range on the dialogue track and flag any clips with excessive noise.
- Measure the true peak audio level on the master and check it meets -1dBTP for streaming.
- Conform the offline edit to the 6K camera originals and flag any relinked mismatches.
- Prep the final timeline for online: check frame rates, retime any mixed-rate clips, and export an AAF.
- Split the long-form cut into social deliverables: 16:9, 9:16, and 1:1 with safe-area reframing.
- Conform the EDL from the offline Avid session to the RED R3D camera originals.
- Validate the final master against the broadcast delivery spec: resolution, loudness, safe area.
- Reframe the 16:9 master for Instagram Reels: vertical crop with subject tracking.
- Create a 4:5 vertical version of the trailer for Instagram feed with blurred pillarboxing.
- Prep the deliverables for three platforms: YouTube 4K, Instagram Reels, and broadcast HD.
- Parse the EDL from the Avid output and verify every cut matches the locked picture.
- Export a clean version and a subtitled version of the final cut for the distributor.
- Export an XML of the current timeline for DaVinci Resolve with all colour metadata.
- Export the current Premiere Pro sequence as FCPXML for the colourist in Resolve.
- Burn timecode into the client review copy and export at 720p with a watermark.
- Render the current Resolve timeline to DNxHR 444 for the online conform.
- Conform the FCPXML from Final Cut to Resolve, preserving transitions and speed ramps.
- Extract the audio from the scratch mix and export it as stems for the sound designer.
- Render a low-res preview of the assembly timeline so I can review it on my phone.
- Generate a delivery package for Netflix: check resolution, codec, loudness, and colour space.
- Export individual stems for dialogue, music, and effects from the final mix.
- Burn subtitles into the screener copy and add the production company logo as a bug.
- Generate a paper edit from the transcript: keep only the answers about funding.
- Find the moment where she says 'that changed everything' and mark it as selects.
- Transcribe all dialogue in the assembly and find every mention of the suspect's name.
- Transcribe the French interview and translate the subtitles to English SRT.
- Find the take where the actor breaks character and laughs. It's somewhere in scene 7.
- OCR the whiteboard in the wide shot and extract the text for the production notes.
- Search the transcript for every question the interviewer asks and list them with timecodes.
- Create chapter markers at each topic change in the interview and export for YouTube.
- Split the 90-minute concert film into individual songs using the setlist.
- Add chapter markers wherever I say 'next step' in the tutorial voice-over.
- Find every clip where the boom is visible in frame and flag it for the editor.
- Extract every frame where the product is visible and export as stills for the client deck.
- Find every shot where the camera is handheld and the horizon isn't level.
- Generate a contact sheet of keyframes from the entire documentary: one frame every 30 seconds.
- Detect all the on-screen graphics and text overlays in the current cut.
- Describe what's happening in each keyframe of the montage sequence.
- Find every clip where someone walks through the frame left to right.
- Cluster the rushes by visual similarity and show me the groups.
- Detect all the scene changes in the documentary and add markers at each transition.
- Create a before-and-after split screen of the VFX shots for the supervisor review.
- Auto-sync all four camera angles for the panel discussion and create a multicam sequence.
- Create a multicam sequence from the three ISO cameras and the switchboard feed.
- Sync the GoPro footage to the main camera using the clap at the start of each take.
- Build a split-screen comparison of the wide and close-up for the client review.
- Ingest the Atomos recorder files and cross-reference timecode with the main camera.
- Ingest the new camera cards, transcode to ProRes, and add them to today's bin.
- Scan the new hard drive for media, ingest everything, and organize by camera and date.
- Create ProRes proxies for the entire media pool so editorial can work offline.
- Organise the rushes by scene and slate number, then create bins for each setup.
- Cluster all the interview footage by speaker and create a separate bin for each person.
- Check the render queue status in Resolve and tell me how long the current job has left.
- Switch to the colour page in Resolve and apply a node tree from the reference clip.
- List every marker on the Resolve timeline and export them as a CSV for the producer.
- Write a Python script to rename all clips using the slate metadata and scene number.
- Write a script to batch-export every timeline as a separate ProRes file.
- Parse the camera reports and flag any clips where the focus puller noted a problem.
- Parse the script breakdown PDF and match each scene to the corresponding rushes.
- Create a skill that automatically preps every new project with standard bins and naming.
- Build a selects reel from the casting tapes: only the callbacks, sorted by role.
- Create a datamosh transition between the two hero shots for the music video.
- Build a string-out of all drone footage sorted by location metadata.
- Find the moment in the ceremony where the couple exchanges vows. It's the wide angle.
- Check if any clips in the project have mismatched frame rates and flag them.
- Build a rough cut from the shot list the director sent over in the PDF.
- Tell me which scenes haven't been covered yet based on the script and what we've shot.
- Find the best establishing shot of the building and use it as the opening frame.
- Identify which clips are missing audio and flag them in the database.
- We've got 14TB of rushes across six drives: ingest everything and build a searchable index overnight.
- Cross-reference 200 hours of archival footage against the paper edit and pull matching clips.
- Conform the locked 8-episode series to 8K ARRIRAW originals and validate every relink.
- Generate proxies for the entire 40TB media library so the remote team can start cutting.
- Deliver the finished series in 23 specs: broadcast HD, UHD, cinema DCP, plus all social cuts.
- Batch-conform all ten episodes from Avid EDLs to the camera originals on the SAN.
- Index the corporate archive, 15 years of video assets, and make it searchable by topic.
- Validate every deliverable in the package against the broadcaster's technical spec sheet.
- Transcode the entire RED shoot to ProRes 4444 for the DI suite and verify frame integrity.
- Render versioned masters, textless, subtitled, and audio-described, for international distribution.
- We shot 400 hours for this documentary. Find every mention of the river and build a thread.
- Build a narrative structure from the interviews: group by theme, not chronology.
- Identify recurring visual motifs across the whole film and create a supercut.
- Pull every emotional reaction from the contributors and rank them by intensity.
- Map every location mentioned in the transcripts and cluster the footage by geography.
- Build a paper edit of the 3-hour assembly: keep only the strongest narrative beats.
- Analyse the pacing of act two: show me the average shot length and flag any scenes that drag.
- Cross-reference the shooting schedule with what we actually shot and flag missing coverage.
- Find the b-roll that best illustrates each interview segment and suggest cutaway placements.
- Transcribe all 60 interviews, identify the 12 key characters, and build a selects reel per person.
- Generate a story map, every scene, every character, every location, from the raw transcripts.
- Find the five most cinematic drone shots from the entire production and sequence them.
- Cut the music video to the beat: place every edit on the downbeat of the chorus.
- Create 15 TikTok cuts from the music video: each one a different visual hook.
- Sync the performance footage to the studio master and comp the best takes.
- Build a vertical edit of the dance sequence with beat-matched cuts for Instagram Reels.
- Datamosh the bridge section and reverse-ramp into the final chorus.
- Extract the hook, 3 seconds max, and loop it as a GIF for the artist's socials.
- Reframe the widescreen music video into 9:16, 1:1, and 4:5 for every platform in one go.
- Find the frame with the most visual energy in the performance and use it as the thumbnail.
- Speed-ramp the skateboard trick: slow-mo on the flip, snap back to full speed on the landing.
- Build a lyric video: burn each line onto the performance footage timed to the vocal.
- Cut a 30-second teaser from the full music video and optimise it for YouTube Shorts.
- Detect every beat and tempo change in the track and place markers on the timeline.
- QC the final master: check for black frames, audio dropouts, and out-of-spec levels.
- Generate a DCP package for theatrical: JPEG 2000, 5.1 surround, XYZ colour space.
- Prep the textless master, the M&E mix, and the dialogue-only stem for the foreign dub.
- Conform the locked picture from the Avid AAF, online in Resolve, and render the broadcast master.
- Check every clip in the timeline for illegal colours and flag anything outside broadcast safe.
- Create a compliance edit: blur faces, bleep profanity, and export a pre-watershed version.
- Sync all five camera angles from the ceremony and build a multicam cut in 20 minutes.
- Find every shot of the couple during the reception and build a 5-minute highlights film.
- Pull the best moments from the speeches: just the laughs and the tears.
- Deliver the ceremony in full plus a 90-second Instagram cut and a vertical story version.
- Ingest a year's worth of product shoots and tag every clip by SKU from the filename.
- Cut 40 product videos from a single shoot day: one per SKU, 15 seconds each.
- Search the corporate archive for every clip featuring the old logo before the rebrand.
- Deliver the brand film in 12 aspect ratios and 4 durations for the media buyer.
- Transcribe the CEO's keynote, pull the three strongest soundbites, and cut them for LinkedIn.
- Find every shot where the actor is holding the red cup.
- My thesis film is due Friday, help me assemble the rough cut.
- I shot on three different phones, make the colour consistent.
- Colour match the iPhone footage to the Alexa so it cuts together.
- I filmed my short on an iPhone in the wrong frame rate, help.
- I accidentally deleted the timeline, can you rebuild it from the EDL?
- Sort 800 clips from three cameras by scene and angle.
- Check all clips for black frames and corrupt files before I start.
- Build a visual mood board timeline from the director's references.
- Generate a quick assembly from the selects bin in scene order.
- Find the take where she nails the monologue on the staircase.
- Cross-reference the script notes with the actual takes we shot.
- Check which takes have clean audio and which need ADR.
- Detect clapper slates in the rushes and auto-sync sound to picture.
- Analyse the pacing of act two, it feels slow.
- Find any continuity errors where the prop switches hands.
- Detect scene changes and auto-split this tape capture into clips.
- Suggest a cut order for the climax based on emotional arc.
- Flag every clip with camera shake worse than our threshold.
- Build a timeline of just the string quartet from the four ISO feeds.
- Cut a 60-second Instagram reel from yesterday's product shoot.
- Sync all the interviews from Tuesday's Zoom call to the B-roll.
- Build a 30-second teaser using only the best crowd reactions.
- Apply our brand LUT to all 47 testimonial clips.
- Create five A/B test versions of the hero ad with different CTAs.
- Batch export the webinar into 12 topic chapters with titles.
- Chop the 90-minute webinar into social clips under 60 seconds each.
- The client wants the logo 10% bigger in all 23 deliverables.
- Find all clips where the talent says the product name wrong.
- Generate a paper edit from the interview transcripts about childhood.
- Repurpose the hero video into 4 platform-specific aspect ratios.
- Build a highlight reel from 6 hours of conference talks.
- Export separate 16x9 and 9x16 versions for YouTube and TikTok.
- Batch apply the client's colour correction notes to all 30 spots.
- Prep the master for Netflix delivery with their spec sheet.
- Create international versions with swapped title cards for 8 markets.
- Transcribe all the Spanish interviews and burn in English subs.
- Extract every music cue and list them with timecodes for licensing.
- Compare the locked cut to the previous version and show me the diffs.
- Build a stringout of every wide shot from the festival stage.
- I need 15-second loops from the best visual moments for the LED wall.
- Find every title card across 200 episodes that uses the old branding.
- Separate dialogue, music, and effects into stems for the mix.
- Pull every mention of the murder weapon from 400 hours of depositions.
- Ingest the 2TB drive from the nature shoot and tag everything by species.
- Find all the archival footage that's 4x3 and flag it for reframing.
- Find the shot where the cheetah starts running in the tall grass.
- Tag every shot in the archive by decade based on visual style.
- Pull every reaction shot of the jury from the courtroom footage.
- Find every usable cutaway from 12 hours of observational footage.
- Find all the establishing shots of London from the 50TB archive.
- Translate the French narration and generate a scratch English dub.
- Cut together every helicopter shot from all three seasons.
- Run Leiden clustering on the whole project to find natural scene groups.
- Identify every speaker in the doc and label their clips.
- Cluster all the talking heads by background so we can group interviews.
- Pull every close-up of hands from the cooking documentary.
- Build a sizzle reel from this season's best moments for upfronts.
- Conform the locked cut from Premiere to Resolve for grading.
- Export dailies for the studio with burned-in timecode and watermarks.
- Pull VFX shots from episode 3 and export as EXR sequences.
- Loudness check all deliverables for EBU R128 before we send to broadcast.
- The broadcast master needs -24 LKFS, fix the mix.
- I need the FCPXML for this Resolve timeline so the editor can finish in Premiere.
- Export clean stems and a printmaster for the cinema DCP.
- Create a CDL for each scene so the colourist can start working.
- Create a Resolve compound clip for each scene from the daily XMLs.
- Reconform the locked picture to the new VFX shots from the facility.
- Detect every beat drop and cut the visuals to match.
- Add datamosh transitions on every cut in the chorus.
- Sync the lyric overlay to the vocal timing across the whole track.
- Sort the BTS footage by setup so I can make a making-of.
- Grade the music video to look like expired 16mm film stock.
- Find every take where the conductor's baton is in frame.
- Make the rap verse cuts land exactly on the snare hits.
- The hip-hop video needs a speed ramp into every verse transition.
- Find the exact frame where the dancer lands the jump.
- I need 20 GIF loops from the electronic set for social promotion.
- Cut every jump cut from my last vlog automatically.
- Find all the moments where I say 'um' or 'uh' and remove them.
- Remove every sponsor segment from my uploads folder.
- Export a 1080p master and a vertical cut for Shorts.
- Chapter this tutorial into sections automatically.
- Match the colour grade to my last episode.
- Find the three funniest moments from the 8-hour stream.
- Find the most viral-worthy 30 seconds of this vlog.
- Pull all my reaction cam clips where my jaw drops.
- Pull every clip where the screen recording shows a terminal window.
- Match my talking-head exposure to the B-roll level.
- Create a multi-language subtitle track for YouTube.
- Cut a stinger for the mid-roll break using existing footage.
- Find every clip where graphics cover the lower third.
- Detect every moment the interview subject goes off-topic.
- Generate a rough cut from the script I just uploaded.
- Find every clip where the presenter looks directly at camera.
- Detect and remove all dead air longer than two seconds.
- Create a vertical 9:16 recut of this horizontal sequence.
- Build chapter markers from the slide transitions in this lecture.
- Find every moment someone holds up a product on camera.
- Detect applause in the keynote and cut away before it dies.
- Find every screen recording segment in this tutorial series.
- Find every clip where there's laughter in the background.
- Datamosh the transition between act one and act two.
- Label which guest is speaking in each segment.
- Build an audiogram for the best 60 seconds of this episode.
- Generate chapter markers with titles for this podcast episode.
- Sync the multi-guest podcast to a single clean mix.
- Auto-label all speakers in this four-guest roundtable.
- Find all clips where dialogue overlaps between two speakers.
- Normalise all dialogue to -14 LUFS.
- Remove background noise from all outdoor interview clips.
- Cluster all interview takes by topic automatically.
- Find the three best interview soundbites under 15 seconds.
- Pull every clip of the red team from the multicam match coverage.
- Run a skills workflow to deliver the full broadcast package.
- Build a string-out of all the confessional booth takes.
- Pull all the frankenbite options for 'I didn't know what to do'.
- Find the act break moments in this reality episode.
- Auto-sort 400 rushes clips into scene and character folders.
- Find the moment she starts crying.
- Build a string-out of every contestant's best moment.
- Log all rushes and flag the best takes automatically.
- Pull every reaction shot where someone laughs.
- Log all the B-roll from today's shoot and sort by subject.
- Find every moment the crowd chants in the concert footage.
- Pull every crowd reaction shot from the concert multicam.
- Find the moment the crowd storms the pitch.
- Pull every SOT mentioning the word 'inflation'.
- Give me all archive clips tagged 'protest' from 2019.
- Cut the breaking news package in under three minutes.
- Pull all archive news packages mentioning this politician.
- Create closed captions for ADA compliance on every export.
- Transcribe and caption all three languages in this interview.
- Export a DPP-compliant delivery package with AS-11 metadata.
- Cut a 90-second highlight reel from today's match.
- Find every clip of player 23 from all six cameras.
- Find the best slow-motion replay of that tackle.
- Build a season recap using only player interview clips.
- Pull every goal scored from all camera angles.
- Pull all the slow-motion shots of the serve from court camera 2.
- Build a highlight reel sorted by loudest crowd reaction.
- Build a two-minute player profile from all season footage.
- Find all clips where the red card was shown.
- Pull every clip where the crowd goes loud.
- Build a multicam sequence from the four keynote cameras.
- Sync all six camera angles to the PA audio.
- Build the day two recap from the conference multicam.
- Build a multi-day event recap under five minutes.
- Build a learning module with chaptered sections and quizzes.
- Detect every hard cut in the rough cut and list the timecodes.
- Apply EBU R128 loudness to every clip before export.
- Send this timeline directly to Premiere with markers intact.
- Apply the same colour science as my previous series.
- Export to DaVinci Resolve with full LUT and CDL baked in.
- Run Leiden clustering on all rushes to find repeating story threads.
- Export a ProRes 4444 master with embedded LUT.
- Find all clips where the interviewee mentions the CEO.
- Ingest this SSD and auto-sort everything by shooting day.
- Pull every moment the host is off-screen.
- Find all b-roll of cityscapes in the archive.
- Find every clip where the subject is backlit and flag it.
- Cluster these 200 b-roll shots by visual similarity.
- Find all clips from the archive shot on an iPhone.
- Pull all archive clips where snow is visible in frame.
- Write a Python script that renames all clips by scene number.
- Detect the beat drops and cut on them.
- Apply the CDL from the DIT to all clips in bin 3.
- Conform my offline edit to the online media.
- Create a virtual tour from drone footage and interior phone clips with auto-sorted locations.
- Burn property addresses into title cards on 50 listing videos at once.
- Match interior lighting across handheld and wide shots in a show home walkthrough.
- Sync ambient audio from a room mic to video recorded on three different devices.
- Tag every drone shot by altitude and speed, then build a highlight reel.
- Organize three days of multi-angle property shoots into folders by room and light condition.
- Grade 30 interior shots to match daylight and tungsten standards across a walkthrough.
- Cut a 90-second listing spot from 2 hours of mixed video, add licensed music, and export.
- Shoot a multi-camera runway show on GoPro and phones, lock multicam timeline in seconds.
- Extract close-up product detail shots from a beauty tutorial and slow them 50%.
- Match skin tone and lighting across before/after clips shot on different phones.
- Tag all fashion footage by garment colour, model, and lighting setup.
- Create a gallery of beauty product close-ups with consistent white balance and exposure.
- Organize 200 lookbook clips by outfit, season, and vibes with one command.
- Build a 15-second product demo with speed ramps, colour grade, and text overlays.
- Extract 30-second interview answers from a 2-hour beauty influencer shoot.
- Overhead camera: automatically detect knife cuts and align footage to recipe narration.
- Speed-ramp a 10-minute cooking sequence into 2 minutes with smooth acceleration curves.
- Match white balance across phone, overhead camera, and finished dish close-up.
- Extract only the moments when the chef is speaking, tag by ingredient, cluster by technique.
- Create three 15-second social cuts from one 3-minute recipe video, each with captions.
- Sync background music to natural beats in a cooking montage: chopping, sizzling, plating.
- Auto-organize 500 food footage clips by dish, cooking stage, and camera angle.
- Build a step-by-step assembly video with slow-motion details and 60fps sizzle shots.
- Merge GoPro, drone, and iPhone footage from a 10-day backpacking trip into one story.
- Auto-sort 1000+ video clips by location, date, and camera model.
- Match colour grade and exposure across morning golden hour, noon, and blue hour shots.
- Extract every wide establishing shot and every close-up reaction from two days of adventure footage.
- Conform GoPro, DJI drone, and iPhone ProRes to DCI 4K 24fps for final edit.
- Cluster travel footage by aesthetic, golden hour, moody blue, vibrant street, and auto-build montages.
- Create a 60-second highlight from 40 hours of mixed travel footage with beat drops.
- Sync action footage to the beat of a licensed track while keeping dialogue intelligible.
- Organize and caption a baby's first year in monthly chapters with auto-extracted highlights.
- Clean up shaky phone footage of a family wedding and grade to match the photographer's style.
- Extract holiday highlights: kids opening presents, cake cutting, everyone laughing.
- Match skin tones across footage shot by three different family members on different phones.
- Build a 3-minute anniversary video from phone footage spanning 10 years.
- Automatically colour-correct a vacation video to remove blue cast from cloudy indoor footage.
- Extract all pet moments, dog jumping, cat napping, and cluster by pet and activity.
- Create a wedding guest highlight reel with music, titles, and captions.
- Sort 200 phone videos from a holiday into folders by family member and location.
- Grade a home video collection to a consistent warm, nostalgic palette.
- Conform MP4, MOV, and vertical phone videos to landscape 1080p for a family compilation.
- Add fade-in, soft background music, and titles to 50 family moment clips.
- Sync a family dinner conversation to soft ambient music without losing any dialogue.
- Create a generative time-lapse from still frames of a 12-hour event with smooth motion.
- Build an abstract video art piece by layering, colour-shifting, and tempo-mapping found footage.
- Detect and enhance glitch artifacts in corrupted video files for experimental compositions.
- Cluster footage by colour and motion energy, auto-create four different montages.
- Conform archival video from DV tape and VHS and grade to modern colour space.
- Sync audio from vinyl, tape, and field recordings to video for an art installation.
- Create a 10-minute gallery loop with seamless audio, no visible cuts.
- Extract every transition and effect moment from 8 hours of experimental footage.
- Process RAW BRAW files: debayer, correct white balance, apply ACES colour science.
- Convert 120fps RED footage to 24fps with optical flow retiming.
- Batch apply complex FFmpeg graph: deinterlace, scale, colourspace, LUT, loudness normalize.
- Parse and apply a CDL colour grade from Baselight to 100 ProRes clips in Resolve.
- Query the database: find all shots where colour saturation exceeds our threshold.
- Regenerate embeddings on a reshuffled media database and rebuild semantic indexes.
- Batch convert 200 MOV files to ProRes 444 and generate thumbnails.
- Measure EBU R128 loudness across 50 clips and auto-normalize with dynamic range preservation.
- From my phone on set: search for all the beach clips we edited today.
- Check the render queue from my phone: how long left on the current job?
- Text from Telegram: create a 15-second highlight from today's selects.
- Quick phone command: apply my daylight grade preset to all selected clips.
- From set: sync this interview audio to the B-roll while I'm away from the suite.
- Send a Telegram clip and ask EditAssist to find every shot of that person.
- Phone check: did the batch conform finish? How many errors?
- Organize footage into folders from my phone and send me a summary.
- Ingest the rushes, detect scenes, conform to project specs, grade, and build a rough cut.
- Multi-camera shoot: sync dailies, conform to spec, build rough assembly, then grade.
- Ingest archival footage, enhance quality, apply colour grade, export to three delivery specs.
- Extract audio, transcribe, generate captions, apply grammar fixes, and export SRT.
- Reorganize a 6-month edit project: re-sort media, regenerate embeddings, rebuild search.
- Group all the drone shots by location using GPS and visual similarity.
- Tag all underwater shots and rate them by clarity.
- Auto-sort the multicam wedding footage by ceremony, speeches, and dance.
- Find every shot with lens flare from the golden hour session.
- Ingest yesterday's cards, proxy everything, and sort by scene.
- I need to submit a festival DCP but I've never made one: walk me through it.
- Can you fix the audio hum on the dialogue? We recorded near an air conditioner.
- My short film is 22 minutes: find the weakest scene and show me what to cut.
- Export a version with subtitles for the deaf jury panel at the festival.
- I shot 16:9 but the gallery screening needs 2.39:1 letterbox: add the mattes.
- We lost the sound from one take: find the same line in another take and patch it in.
- Colour grade my film to look like it was shot on Super 8.
- I need to create a showreel from my three best student projects.
- Help me build a split-screen essay comparing two interview subjects.
- Find every wide shot in my doc that could work as chapter cards.
- Pull the top-performing 5 seconds from each of our 20 ad variants.
- We need an internal comms video from the all-hands: just the CEO's section.
- Batch watermark all client review copies with their company name.
- Create a case study video from three client Zoom testimonials and our product B-roll.
- Pull every product close-up from the launch event and build an Instagram carousel video.
- Recut last quarter's hero video for the new brand guidelines: swap the LUT and end card.
- Generate 30 thumbnail options from the video and pick the frames with the most energy.
- Our agency needs the same spot in 7 languages: swap audio and burn in local subs.
- Find every testimonial clip where the customer mentions ROI or savings.
- Build a year-in-review compilation from every video we published this year.
- Find the funniest fail from my last 10 vlogs and make a compilation.
- Remove every 'like and subscribe' from the last 50 episodes.
- Auto-generate chapters from my talking points and export for YouTube.
- My face cam froze for 30 seconds: cover it with B-roll automatically.
- Find every clip where the game chat gets toxic and bleep it.
- Extract the top 10 reactions from my stream VOD for a highlights video.
- Speed up the boring parts of the tutorial and keep the important bits at normal speed.
- Composite my green screen face cam over the gameplay footage.
- Find every unboxing moment across all my tech review videos.
- Make a best-of compilation from my first 100 episodes.
- Find the moment my guest tells the story about the taxi. It's gold.
- Cut a 60-second teaser from the best argument in today's episode.
- Detect every time I talk over my guest and flag it.
- Create waveform audiograms for the 5 best quotes from this week's episode.
- Sync the remote guest's Zoom audio to the local high-quality recording.
- Find every penalty kick from this season and create a supercut.
- Build a pre-match hype video from training footage and last season's best goals.
- Track player movement across the wide camera and create a heat map overlay.
- Sync all 12 cameras from the boxing match to the referee's microphone.
- Create a slow-motion replay of every three-pointer from all angles.
- Map the emotional arc of the film: find the peaks and valleys in tone.
- Find every archival photo that appears on screen and extract it at full resolution.
- Cross-cut between two interviews to create a dialogue that never actually happened.
- Find every mention of the war across 30 hours of oral history recordings.
- Build a timeline of events from the transcripts: chronological, not interview order.
- Generate versioned masters for 15 international territories with local legal bumpers.
- Process three weeks of dailies overnight: proxy, sync, sort, and build selects.
- Check every deliverable against the studio's technical QC spec before upload.
- Build an IMF package with supplemental audio tracks for the streaming platform.
- Conform the VFX pulls back into the online timeline with handles.
- Find every shot of the grandmother. She passed away last month and the family wants a tribute.
- Sync the DJ's feed to the dance floor multicam.
- Build a same-day edit from the ceremony footage. It needs to play at the reception.
- Generate a twilight composite: blend the blue hour exterior with the warm interior.
- Cut 50 property videos in one session: same template, different footage and titles.
- Find the money shot, the moment the cheese pulls perfectly, and loop it for Instagram.
- Sync the ASMR mic to the overhead camera for the cooking montage.
- Find every sunset from the three-week trip and build a 30-second supercut.
- Stabilise the handheld hiking footage and match it to the steady drone shots.
- I just want to trim the boring bits from my holiday video. Keep it simple.
- Merge my wife's phone clips and mine into one video sorted by time.
- Our daughter's school play: sync the two parent phone recordings into one cut.
- Reverse the entire film and re-score it to a drone soundtrack.
- Create a visual rhythm: cut every shot to exactly 1.5 seconds for a hypnotic montage.
- Layer four different angles on top of each other at 25% opacity for the art piece.
- Use the narrative director to analyse the full corpus and suggest a story structure.
- Run multimodal search: find clips that look like this reference frame AND mention the bridge.
- Set up a skill that auto-delivers every new project in our standard 8-spec package.
- Apply blend modes to the overlay track: screen for the light leaks, multiply for the shadows.
- Batch edit the timeline: add 12 clips, set opacity on 3, split 2, reorder track 4: one command.
- Calculate TC offsets for the multicam sync using SMPTE timecode from the camera metadata.
- Check title safety on the lower thirds across all 32 delivery specs.
- Analyse the waveform: show me RMS, true peak, crest factor, and dynamic range.
- Detect every beat, tempo change, key change, and section boundary in the music track.
- Search by voice: find every clip where this specific person is speaking.
- Grade the entire short film to look like Kodak Vision3 500T.
- Find every take where the wind noise ruins the dialogue.
- Pull every reaction from the audience cam during the live taping.
- Create a 30-second promo cut with captions for LinkedIn and one without for the website.
- Split the interview into individual answers and file each one by topic.
- Conform the 4K master down to HD and check nothing gets cropped.
- Find the five longest continuous takes in the whole project.
- Detect every shot where natural light changes mid-take and flag for the colourist.
- Sync the BTS photographer's audio recorder to the main camera for the making-of.
- Build a mood reel of textures and details, hands, surfaces, fabrics, from the shoot.
- Create a recap bumper for the top of episode 4 using clips from episodes 1 through 3.
- Export frame grabs from every keyframe in the colour-graded timeline for the lookbook.
- Group all outdoor shots by weather condition: sunny, overcast, rain, fog.
- Find every time the director yells 'cut' in the room tone and use it to split takes.
- Build a pre-vis animatic from the storyboard scans with rough timing.
- Create a censored version: blur licence plates, bleep surnames, remove identifying marks.
- Pull the cleanest 10 seconds of room tone from each location for the sound designer.
- Find every shot that matches the aspect ratio of the reference film I uploaded.
- Detect all the Dutch angles in the thriller and list them with timecodes.
- Auto-generate a shot list from the finished edit: shot type, duration, description.
- Build a dialogue-free version of the trailer using only music and visuals.
- Convert the paper edit into a Resolve timeline with in and out points.
- Parse the scene breakdown and auto-populate the media pool with matching takes.
- Use the shooting script to find the best matching take for every line of dialogue.
- Build an assembly from the script notes PDF and lay takes in scene order.
- Auto-generate a text-based edit from the interview transcript, Descript-style.
- Grade the entire sequence to match the look of Blade Runner 2049.
- Apply a teal and orange grade to the action sequence and generate a LUT.
- Extract the colour science from this reference frame and apply it across the timeline.
- Match all six camera angles to the A-cam hero grade so they cut seamlessly.
- Grade the LOG footage to Rec.2020 PQ for HDR delivery with a DolbyVision trim pass.
- Auto white-balance every clip in the media pool before the colourist starts.
- Remove background noise from the outdoor interview clips and boost dialogue clarity.
- Detect and suppress the air conditioning hum across all interior interview clips.
- Isolate dialogue from the crowd noise in this live event recording.
- De-reverb the voiceover recorded in the tiled room and match it to the studio tone.
- Identify which interview clips need ADR and flag them for the director.
- Find all clips where the audio peaks above 0dBFS and duck those sections.
- Sync the Zoom H6 audio to every clip from today's shoot using waveform matching.
- Auto-sync the lav tracks to the boom for every scene and flag drift beyond 3 frames.
- Sync music video performance audio to the studio master across all eight cameras.
- Generate ProRes LT proxies for the entire RED shoot and link them to the originals.
- Transcode the BRAW files to DNxHR SQ for the Avid offline and verify frame integrity.
- Create quarter-res H.265 proxies for the remote editorial team and upload to the NAS.
- Swap from proxy to full-res media and verify every clip relinks correctly.
- Roundtrip the Avid Media Composer timeline to Resolve and back without losing metadata.
- Convert the Final Cut Pro X project to a Premiere sequence, preserving speed ramps.
- Move the locked Premiere edit to Resolve for the final grade with all markers intact.
- Export the colour-graded Resolve timeline back to Premiere with all node data baked in.
- Auto-zoom on the speaker when they gesture or lean forward in the talking-head footage.
- Detect and remove all awkward pauses longer than half a second from the talking head.
- Add auto-captions with word-level highlighting and export for TikTok.
- Detect every face in the video and blur the ones I haven't cleared for release.
- Auto-cut the 4-hour Twitch VOD into the best 10 highlight clips for YouTube.
- Remove the green screen from my face cam and composite over the gameplay at 60fps.
- Detect every sponsor read by voice and segment them with markers.
- Check every clip in the project for variable frame rate and flag the offenders.
- Verify all assets in the project are online and nothing is missing or offline.
- Render the final timeline on the GPU and report estimated time for each codec option.
- Add 24-frame handles to every cut point before exporting the AAF for sound.
- Strip all effects and grades from the timeline and export a clean AAF for the mixer.
- Watch all the takes and star-rate them 1 to 5 based on focus, framing, and performance.
- Auto-flag every clip where the boom dips into frame or a crew member is visible.
- Rate the B-roll by visual quality and emotional impact, then sort by score.
- Detect which takes have the best audio SNR and mark them as selects.
- Export markers and notes as a PDF report for the director to review on their phone.
- Generate a change list between the current and previous cut for the producer.
- Export the timeline as an interactive web review page with timecoded comments.
- Prep the podcast episode for Spotify: chapters, cover art per segment, and LUFS normalised.
- Cut the worship service into individual sermon, song, and announcement segments.
- Process all the body cam footage, redact faces automatically, and export with chain of custody metadata.
- Build a medical training video from the surgical footage with annotations at key moments.
- Cut the deposition video by question and answer, with timecoded transcript overlay.
- Use scene detection to auto-split this tape capture into individual clips.
- Cluster all footage by visual mood: warm, cold, neutral, dramatic, soft.
- Generate a continuity report from the footage: flag mismatched props, wardrobe, and positions.
- Predict the optimal cut points in this interview based on sentence endings and breath pauses.
- Auto-generate B-roll placement suggestions based on what the speaker is discussing.
- Detect duplicate or near-duplicate clips across the entire media library.
- Analyse shot composition across the sequence and flag any that break the 180-degree rule.
- Prep today's dailies: ingest cards, generate H.265 proxies, sync dual-system audio, and deliver a log to the producer.
- Run the online conform: relink the offline Avid cut to the 4K ARRIRAW originals, flag missing handles, export the relink report.
- Pull VFX shots from the locked picture: export EXR sequences at 4K with 8-frame handles and a CSV pull list.
- Build the colour pipeline: apply DIT CDLs to dailies, export a graded XML for the colourist, and prepare the LUT folder.
- Deliverables prep: 4K broadcast master, 1080p streaming, stereo mix-down, and a 9:16 social cut. One batch.
- Prep the OMF for the sound house: strip to dialogue and effects, add 10-frame handles on every cut, export with embedded audio.
- Set up the grade session: import the FCPXML from editorial, apply the show LUT, create a CDL node for each scene.
- Build the international delivery package: swap M and E stems, burn in localised subs for 6 territories, QC loudness on every version.
- I don't know how to export from Resolve for YouTube. Walk me through the codec and bitrate settings.
- Help me understand what an AAF is and how to send one to my sound designer.
- What's the best way to organise my media pool before I start a new project?
- I'm new to colour grading. Show me how to read the scopes and set levels correctly.
- How do I sync dual-system audio if I forgot to use a clapperboard?
- Help me cut my first short film. 3 hours of footage and no idea where to start.
- Find every shot of my dad from the family reunion. He passed this year and we want to make a tribute.
- Cut together my daughter's first year. Every milestone: first steps, first laugh, first birthday.
- Find the moment the bride's father sees her for the first time in her dress.
- Build a memorial tribute from 30 years of home video. Group by decade.
- Pull every moment the kids laugh in the holiday footage. Just the laughs.
- This documentary took 6 years. Help me find the single most powerful 90 seconds in the whole film.
- Clip the best 60 seconds from this podcast for a LinkedIn video with auto-generated quote cards.
- Create an audiogram: waveform animation, speaker photo, and the best 45-second quote for Twitter.
- Cut 5 YouTube Shorts from this 20-minute tutorial. Each one self-contained with a hook in the first 2 seconds.
- Turn this horizontal interview into a TikTok with dynamic cropping that follows whoever is speaking.
- Build an Instagram carousel video from this conference talk: 10 slides, one key point each.
- Clip highlights from a 6-hour Twitch stream. Find the five moments with the biggest chat reaction.
- Parse this CMX 3600 EDL and verify every event against the source media in the bin.
- Generate an ACES config for the project: set the working space to ACEScg, input transforms per camera, and output to P3-D65.
- Build an IMF package with supplemental audio tracks and a CPL for the streaming platform.
- Export a complete DCP: JPEG 2000 MXF with 5.1 audio, XYZ colour space, and KDM generation info.
- Reconform the locked picture after VFX turnover: relink updated shots, verify handles, rebuild the render queue.
- Find the exact frame where the cat knocks the glass off the table.
- Build a supercut of every time someone says 'actually' in the entire series.
- Find the frame where the goalkeeper's face says everything. You'll know it when you see it.
- Make a compilation of every awkward silence longer than 4 seconds in this reality show.
- Find every clip where a dog appears in the background of a scene it wasn't supposed to be in.
- Create a supercut of doors closing. Every door. In the whole film.
++
FeaturesAn agent with 150 tools,
An agent with 150 tools,
operating your edit suite
01
Agentic NLE Control
Direct control of DaVinci Resolve, Premiere Pro, and After Effects. The agent reads your timeline, executes edits, applies grades, and triggers renders. Autonomously.
02
Semantic Search
Every frame indexed. Every word transcribed. Describe what you're looking for and the agent finds it. Across terabytes, in seconds.
03
Autonomous Assembly
The agent builds timelines from intent. Rough cuts from transcripts. Selects from director notes. Multi-cam sync from scratch. You describe the cut. It builds it.
04
Specialist Delegation
Nine domain specialists: colour, audio, conform, narrative, ingestion. Complex tasks route to the right agent with the right tools. One command, many minds.
++
++
Architecture
How it works
One binary. No cloud processing of your footage. Everything runs on your machine.
01
The agent operates your timeline
Not a suggestion engine. Not a chatbot with opinions. EditAssist takes direct control of your NLE: reading the timeline, executing edits, applying grades, triggering renders.
Nine specialist agents handle domain-specific work: colour grading, audio mixing, conformance, narrative structure, ingestion. The orchestrator delegates. The specialists execute.
DaVinci Resolve
Live plugin: read/write timeline, grade, render
Premiere Pro
CEP panel: live metadata sync, edit operations
After Effects
CEP panel: composition control, render queue
02
Connect everything you use
EditAssist connects to your existing workflow via the Model Context Protocol (MCP). Meeting notes, task boards, review platforms, spreadsheets. Context flows in. Decisions flow out.
External tools plug in as first-class data sources. Use the connectors that ship with EditAssist, or write your own. One open protocol, every integration.
03
One instance, every interface
A single agent running on your machine. Access it from the desktop app, a Telegram message, the terminal, or directly inside Resolve and Premiere. Same context. Same memory. Same project state.
Send a command from your phone on set. Check the result from the edit suite. The agent doesn't care how you reach it. It cares about your project.
For programmatic control, connect via WebSocket or run in headless mode. Build your own automations on top of the agent.
04
Your footage never leaves your machine
Transcription, vision analysis, embeddings, search indexing. All of it runs on your GPU. No footage uploaded. No frames sent to a server. No exceptions.
The only data that leaves your machine is the text of your conversation with the AI. Your media, your project files, your client's footage stays on your hardware.
On your machine
TranscriptionVision analysisEmbeddingsSearch indexMedia filesProject state
Leaves your machine
Chat text only
++
++
Now Accepting Applicants
Join the closed beta
EditAssist is free during beta. Applications are reviewed individually. Limited spots available.
++