Dashboard for Swastik

Client: Swastik Stories Analytics Partner: Oogwai My role: UI Designer Platform: Web Dashboard Domain: YouTube Content Intelligence · AI Overview

How might we turn YouTube data and editorial knowledge into a workflow that helps a large content team make better publishing decisions?

I designed a content-operations dashboard for Swastik Stories, a large Indian storytelling and entertainment brand, as part of its work with Oogwai, an AI-powered analytics company.

Swastik Stories is a cultural storytelling company built around mythology, history and entertainment. The company says it has produced 5,000+ episodes, 50+ TV & OTT shows and 25,000+ hours of content, while expanding its distribution across digital platforms and FAST channels.

Oogwai is an AI/data analytics company building products that sit on top of business data and answer questions in natural language rather than simply presenting another static dashboard. Oogwai's YouTube product is positioned around channel intelligence: it brings together retention, CTR, publishing cadence, topic gaps, competitor benchmarks and packaging insights, then translates those signals into recommendations teams can act on.

Traditional analytics dashboards answer:

But a content team needs to answer:

For example:

Which title is most likely to work?

What description best fits this story?

Which thumbnail should we use?

Which character or moment should become the visual focus?

The interface therefore needed to bridge analytics → editorial judgment → production.

I built a workflow where AI could assist the team while keeping them in full control and designed the dashboard specifically for production tasks rather than filling screens with heavy analytics. “The AI suggests. The editor decides” This is important for a brand like Swastik Stories, where context and storytelling choices matter as much as performance metrics.

Key workflows are visible in the current product:

A structured generation interface that gives the AI enough editorial context to produce relevant candidates.

The interface allows the editor to request multiple candidates.

That changes the relationship between AI and editor.

Instead of:

The workflow becomes:

The AI becomes a creative collaborator, not an autonomous publishing system.

An image-editing workflow designed around controlled changes, rather than completely regenerating a design.

The thumbnail workflow follows a completely different model.

For a professional content team, accidental changes to the original design can be costly.

So the product establishes a clear contract: Preserve the original → Change only the requested element.

The History screen creates a record of everything generated inside the platform.

While Title & Description and Thumbnails focus on creating, History focuses on reviewing and retrieving.

The interface separates previous work into two workflows:

Writing runs

Thumbnail sessions

This makes it easy for editors to understand what has already been generated and return to previous work when needed.

The Settings screen is where the product becomes personalized to the organization's actual content operation.

Rather than treating settings as technical preferences, I designed them around the things that influence everyday content generation:

Saved links

Shows

Pillars

Tiers

Taste profile

Image-generation model

This means the system can learn the organization's working language and creative preferences.

Looking at the complete experience, each screen solves a different part of the content workflow.

Together, these aren't isolated features.

They form a connected workflow:

The core design challenge wasn't simply:

It was:

The solution was to build a system where context, generation, review, and operations work together.

The AI suggests. The editor decides. The system remembers. The team scales.