All posts

Five Tutorials, Two Prompts Each: The Tellers Series Is the Showcase

We just published Tellers Basics, a five-part tutorial series on vibe editing. Each episode shows something you can do with Tellers. But the finished videos also show how we made them: two prompts per episode, a consistent editing style, and about two hours to edit the series.

I started with several hours of recordings, separate microphones, and separate screen captures. I gave Tellers a link to the rushes, described the edit, and closed my computer. Tellers downloaded and analyzed the material in the cloud, synchronized the audio, removed silence and hesitations, and added motion design.

The tutorial series itself is the real Tellers showcase. You can watch the lessons to learn the product, and watch the editing to see what the product made.

Watch the five-part Tellers Basics series

Start with the full Tellers Basics playlist, or jump to the workflow you want to learn:

  1. Simple Stock Footage Edits — start with a simple edit using stock footage.
  2. Cut Your Own Footage With AI — work with your own recordings.
  3. Create Motion Graphics on Your Videos With a Simple Prompt — add motion design through natural-language instructions.
  4. All Generative AI Models at Your Fingertips — explore the generative tools available in Tellers.
  5. Create Specialised Assistants to Automate Specific Formats (+ Pro Tips) — build repeatable workflows for formats you make regularly.

The recordings were stored on my NAS, short for network-attached storage. Think of it as a box of drives connected to a network: it holds your files independently of your laptop and can make selected folders available through a share link.

In video production, rushes means the raw, unedited recordings. Mine included the presenter footage, the screen recordings, and the higher-quality audio recorded on separate microphones. Before the creative edit could happen, those sources needed to be downloaded, understood, and aligned.

I pasted a public share link to the NAS folder into my prompt. Tellers fetched the files and analyzed the videos in the cloud. I could close my computer after sending the request because the work was running on Tellers’ servers. The NAS still needed to stay online and the link reachable during the download; my laptop did not have to do the processing.

That changed the starting point of the project. I could give Tellers the source material and the brief in the same conversation, then let it work through the footage.

The first editing prompt

The creative direction was straightforward: keep me visible alongside the demo, introduce me, give the overlays a distinctive look, tighten the pacing, and make each episode work on its own.

Here is the editing brief:

Keep the presenter in picture-in-picture most of the time.

Add an on-screen introduction for Robin Guignard-Perret, co-founder and CEO of Tellers. Give the overlays a distinctive, stylish, brutalist look.

Cut hesitations and silence where needed to keep the pacing tight.

Add gentle background music.

Remove references to other tutorial episodes so this video works as a self-contained lesson.

Use a splash screen and sound similar to the reference timeline: [Thousands of Hours - TikTok 20s].

PiP means picture-in-picture: the presenter stays visible in a smaller frame while the screen recording carries the demonstration. The reference timeline gave Tellers an existing direction for the splash screen and sound.

The brief combined editorial decisions with visual direction. Cutting hesitations sets the pacing. Making the lesson self-contained tells the agent which digressions to avoid. A stylish, brutalist look gives the overlays a visual identity. You can learn more about making that direction precise in our guide to better AI video prompts.

To adapt the approach, open a tutorial editing prompt in Tellers. Add your own recordings and reference material to the conversation.

Two prompts per episode, one style across the series

Each of the five episodes was edited with two prompts. Across the series, Tellers handled the separate sources, synchronized the microphone audio, removed silence and filler words, and added motion design. The episodes kept a consistent visual style and editing treatment.

The edit took about two hours for the series. The several hours of recording were the source material for that edit.

That consistency matters for tutorials. The presenter, screen capture, overlays, and pacing need to feel like parts of the same series, while each lesson remains useful to someone who lands on it first. The prompt made that intention explicit: avoid references to other episodes and make each video self-contained.

This is the workflow we describe in our articles about working with existing training footage and Tellers’ cloud video infrastructure. Here, you can see the finished result in the tutorials themselves.

One more prompt for a batch of social shorts

The same rushes also became material for social posts. I created 12 shorts with a single prompt, asking Tellers to find moments that either demonstrate the product’s value or have the potential to catch attention on social media.

Here are two of the shorts created from those recordings:

The brief covered the whole batch: clip selection, audio synchronization, hooks, and a separate timeline for each short.

[Your tutorial rushes folder]

Create 12 shorts for Tellers' social channels from the rushes in this folder, each in a separate timeline.

Synchronize the high-quality microphone recording with the face-camera footage and use it to replace the camera audio.

Give each short a strong opening hook: a comment reply, a visual hook, a bold on-screen title, a question, or another creative approach.

Format the videos for Instagram Reels and TikTok.

Be creative. Choose clips that clearly demonstrate what Tellers can do or have strong potential to catch attention and be shared.

Look up current best practices online if needed.

The audio instruction is a small line with a big effect: the face camera’s audio needed to be replaced with the separate, high-quality microphone recording and realigned. That requirement carried into the shorts alongside the creative brief for hooks and clip selection.

The prompt also asked for a separate timeline for each short. That gives each clip its own edit to review and refine, while reusing the recordings already in Tellers.

Try the social shorts prompt in Tellers, then provide your own source folder.

Watch the lesson, look at the edit

When you watch Tellers Basics, look at how the presenter sits alongside the screen recording, how the pauses are cut, how the graphics support the explanation, and how the style carries from one episode to the next. Those are all parts of the production brief Tellers worked from.

Watch Tellers Basics on YouTube, then start your own tutorial edit in Tellers.

Frequently asked questions

How was the Tellers Basics tutorial series edited?

The five episodes were edited with Tellers using two prompts per episode. The source material included several hours of recording, separate microphone audio, and separate screen recordings. The edit took about two hours for the series, with a consistent visual style across episodes.

What is a NAS, and how did Tellers access the recordings?

A NAS is network-attached storage: a device that stores files and makes them accessible over a network. For this production, Robin provided a public share link to the folder containing the raw recordings. Tellers downloaded and analyzed the material in the cloud.

Did the computer need to stay open while Tellers worked?

No. Robin pasted the prompt with the source link and closed his computer. Tellers continued processing in the cloud. The NAS and its share link needed to remain reachable while the files were being downloaded.

Were the social shorts created from the same recordings?

Yes. A single prompt turned the same tutorial rushes into 12 shorts for Instagram and TikTok, each with its own timeline and hook. Tellers also realigned the separate microphone recording and used it to replace the face-camera audio.