Brutalist AI Video Production & Content Pipeline Progress | Akshay Chavan

This is a weekly progress report from Akshay Chavan, a fellow at Humanitarians AI. The week followed a clear arc: learn the Brutalist video production workflow, contribute to the existing content pipeline, then use the system to build something new. The video walks through each of those steps in turn.

Learning the Brutalist workflow

Akshay started by going through Professor Brown's Brutalist AI video production material. The aim was to understand how Humanitarians AI videos are actually built, from scripting and narration through visuals, rendering, and quality control. The video presents this as the move from learning the workflow to working inside it.

Reviewing and publishing 14 videos

The second step was practical. Akshay worked through the Claude Bear video collection, reviewing and uploading 14 videos. The topics he lists are Claude basics, flashcards, form generation, supplier selection, reorder policies, and workflow automation.

Those videos were published into the appropriate Humanitarians AI YouTube playlists, including Claude Basics and Extending Claude. According to the video, this gave him hands-on experience with the existing content workflow while helping move completed material into the published library.

Building a Kafka explainer

With that experience behind him, Akshay applied what he had learned and created a technical explainer called "How Kafka handles millions of events." The video says the explainer covers these parts of Kafka:

• the durable log
• partitions
• brokers
• consumer groups
• offsets
• rebalancing
• replication

Landscape and portrait, built natively

He produced the explainer through the Brutalist workflow in both landscape and portrait formats. Each orientation uses a native 4K layout built for that orientation. The portrait version is not simply a crop of the landscape one.

The week's progression

Akshay sums up the week in three steps: learn the production system, contribute to the existing pipeline, and then use that system to build something new. The video presents the 14-video review as the bridge between studying the workflow and making an original explainer.

Key takeaways

• The week moved from learning the workflow, to contributing to the pipeline, to building an original piece.
• Professor Brown's Brutalist material covers how videos are built from scripting and narration to visuals, rendering, and quality control.
• Akshay reviewed and uploaded 14 Claude Bear collection videos, from Claude basics to workflow automation.
• The videos went into the Humanitarians AI playlists, including Claude Basics and Extending Claude.
• His own explainer, "How Kafka handles millions of events," covers the durable log, partitions, brokers, consumer groups, offsets, rebalancing, and replication.
• The explainer exists in landscape and portrait, each with its own native 4K layout rather than a crop.
• Contributing to a real pipeline came before building something new with the same tools.

Try it yourself

If you are learning a production workflow, follow the same order the video describes. Study how the existing work is built, help move finished pieces through the pipeline, and only then make something of your own. For a technical topic, you could pick one system, such as the Kafka parts listed above, and plan how its layout would differ in landscape and portrait before you build either.