The New Writer Module

7/17/2024

We've released a big update to our Writer module (previously called the Blog module).

More than one year of learning and feedback have made v2 a significant improvement.

A brief recap of how this module works -

Upload a video into Parmonic ➡️ our engine identifies topics and subtopics based on content analysis ➡️user can choose which topics to include/exclude ➡️ our engine writes the blog ➡️ user can add video moments ➡️ user can download a copy for further editing or sharing.

The big things we learned from users and our research:

1. Most brands have their own style - the structure and layout they prefer. Some prefer paragraphs of text mixed with video moments, some prefer a mix of text and bullet points, some prefer a mix of short and long sections...the list goes on.
2. It's very important to do inline editing and styling so you can finish as much of the draft as possible in Parmonic itself.
3. AI (and LLMs in general) perform better when given more specific instructions about certain variables (e.g. tone, emphasis, etc.)
4. Creating custom prompts and templates in tools like ChatGPT takes a lot of time, energy and skills.

The new Writer addresses these and gives you powerful capabilities.

Now you can -

1. Create your own styles and templates - you can upload past blogs and our engine will decipher your style. You can create multiple styles and templates for various use cases (e.g. webinar, interview, internal content, etc.)
2. You can do inline editing - format text, insert content.
3. You can write specific instructions and prompts - both at an individual video level or as part of your template. (e.g. Emphasize this not that, use this tone, etc.)
4. Our engine will also auto-detect the industry that the content is relevant to and make analysis adjustments on its own.
5. Move things around - you can reorganize sections by dragging them.
6. Automatically get a visually-distinct sub-headline that you can use as the meta description.
7. Rewrite the whole or portions easily.

For this module, we use a heavily-customized LLM under the hood so our users can get a better experience compared to other options. We'd love to hear your feedback.