Seven days. Sixteen hours of hands-on work, in your office. Your team leaves with one way of working they can repeat. Topic, script, storyboard, anchored frames, generation, edit) that plugs into the content pipeline you already run and survives every model upgrade.
Brands and channels we’ve worked with — across channel management, production, campaigns and collaborations. Subscriber counts belong to the channels, not to us.
Heartfulness
Pansari Group
Pot and Bloom
Home Banao — a channel we own
Atomberg
Panasonic India
Anmol Garg
Therrapie Healthcare
Sanjay Sehgal (SehgalNotes)
Bosch India
FanzartMost teams have already tried. Someone opened Veo or Runway. They typed a paragraph. They got eight seconds of something almost right. They tried again and got something else. The slide at the next meeting said "AI video: promising, but not consistent." That is a fair verdict. It is what happens when you prompt, hope and re-roll. It says nothing about what the tools can do when there is a process around them.
The way that works is simple, but you must stick to it. Pick the topic first. Write the script. Draw the shots. Turn those drawings into images. Then fix every shot with a first frame and a last frame. Add a middle frame if the cut needs one. Then prompt only the movement in between. Generation stops being a lottery because you have already decided what the shot looks like at both ends. The edit stops being a rescue and becomes assembly. And because the boards and frames are the asset, the workflow is model-agnostic — plug it into any API, and when a better model ships next quarter you rerun the same boards instead of starting over.
This course puts that method into your team in seven days. It is the one we use ourselves. It is not a tour of tools. It does not depend on one vendor still being the best one in six months.
Six steps, in this order, every time. Steps one to four are the thinking. Step five is where you take control. Step six turns it into a finished video.
Decide this before you open any tool. Who is it for? What one thing must it say? What kind of video is it? And where it publishes. Written down as a brief your team already recognises.
The script comes out as a shot list, not paragraphs. What is said. What is seen. How long each bit runs. Your brand voice and claim rules are applied here, where they are cheap to enforce.
Shot by shot, on a whiteboard or a doc. The framing. The camera move. How long it runs. How one shot leads to the next. This is where a video person earns their pay. It is also where most AI-only teams skip straight past.
Each drawing becomes an image. Same faces, same products, same places, held steady by reference images. Your bosses approve these images here, before a second of video is generated.
You set the first frame and the last frame. You add a middle frame if the cut needs one. Then you say what happens between them. You pick a model per shot for what that shot needs. Batches run through the UI or the API.
Edit, voice, music, colour, brand packaging and the AI label. Check it against the drawings. Then publish it your normal way. Because the ends of every shot were known, the cuts land.
You decide the solid frames. The model only fills the dashed ones. That is the whole trick.
Your team decides the framing, the flow and how the product looks. The model does not get to guess.
Shots cut together because both ends were planned. Editors assemble and grade instead of hunting for a usable second.
The drawings and frames are what you own. Point them at any model. When a better one arrives, run the same drawings again.
Approvals happen early, on the drawings. Legal and brand sign off there, not on a finished video nobody wants to redo.
The same method works for a quick phone clip and for a brand film. Only the planning per shot changes.
Quick product clips for Shorts, Reels and paid social. Fast to make. Many versions. One template reused across products and markets.
A real presenter or voice-over, with AI b-roll behind it. Diagrams that move. Scenes that used to need a shoot.
Hero shots and camera moves around a product that stays the same in every frame. Tested as drawings before you spend money.
Longer sequences that flow, with real camera language and colour work, built shot by shot. This is where the video specialist matters most.
Three specialists work with your team. One for AI models and prompts. One for editing, sound and packaging. One for framing, camera language and what to shoot rather than generate. Every session runs on real briefs from your calendar.
Session 1 (4 hrs): freezing topics, scripting as shot lists, visualising boards on your own briefs.
Session 2 (4 hrs): image models, character and product consistency, style locks, designing first, last and transition frames.
Session 3, four hours. Prompting the movement between frames. Picking a model per shot. Running batches.
Session 4 (4 hrs): assembly, voice and music, grade, brand packaging, disclosure, QA against the boards, publishing.
We look at how work moves through your team today. We pick the real jobs the sessions will use. We choose the tools and set up accounts before anyone sits down to learn.
Four sessions, one a day, in your office or live online. Every session ends with something made, not just explained. By session four your team has finished work built end to end with the workflow.
One job, start to finish, run by your team. We watch and comment. Gaps show up here, and we fix them while we are still there.
You keep the written method, the prompt and frame library you built, the tool list with reasons, and a plan for what to make first.
The price depends on group size, on-site or online, the videos you make and the languages you use. So there is no fixed rate here. Talk to us and we come back with a scoped proposal and the day-by-day outline. No lock-ins, and everything built in the sessions is yours.
We are two partners in Bengaluru who run video and YouTube every week. We have worked on channels like Heartfulness and Pot and Bloom. We own and run Home Banao ourselves.
The method we teach is the one we use. We built it to survive new models, because new models arrive every month.
You can see how we think before you commit. We publish free teardowns of real channels on our teardown channel. Or ask for a teardown of your channel.
Not sure what to make yet? Buy an hour of help first. Need a room to film in? See studio setup.
Tell us what you make, who is on the team and when you are free. Nikhil plans it with you on one call.
Talk to usTell us about your team and the video you need to make. You get a plan in one working day. No sales deck.
Prefer to pick a time? Book a call directly.
