A collaboration between three designers

This project began as a collaboration between three designers—each from different disciplines. I specialize in interaction design and generative technology. Divya is a trained Kuchipudi dancer and designer. Sofia brought expertise in spatial and exhibition design. Our shared goal:

What would it look like to fuse our disciplines to reimagine how people experience cultural performance?

We chose to focus on Kuchipudi, a classical Indian dance form that is rich in symbolism, emotional expression (rasa), and devotional storytelling. For Divya, it was a way to honor her cultural roots. For the rest of us, it became an opportunity to reinterpret tradition through immersive, audience-centered design.

But we also recognized a challenge.

Many people (especially non-Indian audiences) found Kuchipudi visually beautiful but hard to follow. The gestures, lyrics, and symbols are unfamiliar. We started hearing things like:

“I loved the movement, but I didn’t understand the meaning.”

That disconnect became our entry point into the problem space.

Research Activities

To better understand our audience, we conducted:

• 10 semi-structured interviews with past attendees of cultural events

• 25+ survey responses from students, educators, and design peers

• Assumption mapping to identify cultural gaps and anxieties around tech integration

• Secondary research into museum exhibition design and intercultural performance studies

Key Insights:

• Viewers often felt like outsiders. People were ****drawn to the performance but lacking cultural context

• No translations or interpretive guides made it difficult to track story progression

• Long performance durations made it harder for newer audiences to stay engaged

• Some expressed skepticism about using technology in sacred or traditional settings

This research validated what we felt early on: audiences were open and curious, but unequipped. They didn’t need the tradition altered—they just needed a way in.

Defining the Project Goals

Our research pointed to a clear and recurring theme: audiences weren’t uninterested in cultural performance, they just didn’t have a way to access its meaning. The problem wasn’t the tradition, it was the lack of context.

We synthesized our interviews, survey data, and assumption maps into a focused design challenge:

How might we use immersive design to preserve tradition while making its meaning more accessible to new audiences?

We framed our direction around three experience principles:

1. Cultural Respect

   The performance should never feel diluted or simplified. Technology would exist in service of the story, not as a spectacle.

2. Layered Understanding

   Rather than over-explaining, the experience should offer multiple entry points. Audiences should be able to engage at their own pace: visually, narratively, or emotionally.

3. Emotional Clarity

   The design should help unfamiliar audiences feel something, not just observe. If they didn’t understand every symbol, they should still walk away with an emotional connection.

We aligned as a team around this north star: build a performance environment that feels inviting without feeling reductive. A space that welcomes curiosity, supports interpretation, and respects the art.

Develop

Prototyping Cultural Connection Through Visual Design and AI

With our design challenge in place, we began exploring how to bring the story of Alamelumanga to life in a way that honored tradition while expanding accessibility. This stage focused on collaborative ideation, prototyping, and technical experimentation.

Designing a Multi-layered Experience

We decided to build an experience that gave audiences multiple ways to connect. Our early concepts included a looping slide deck, printed narrative guides, and altar-based projections. We wanted these elements to offer just enough context before the performance began, setting the stage without overloading the viewer.

Alongside these exhibition tools, we began experimenting with how artificial intelligence could be used during the live dance performance. This was not about spectacle. It was about using technology to translate the symbolic meaning of the performance into visuals that could guide the audience emotionally.

Generative AI System

As the technical and interaction designer, I led the development of a real-time visual system that would respond to movement and music. This involved:

• Mapping each line of the Telugu lyrics to a visual prompt in English

• Collaborating closely with Divya to ensure visual metaphors stayed true to the choreography

• Integrating MediaPipe to capture motion data from the dancer

• Using TouchDesigner to build a control interface

• Feeding prompts into StreamDiffusion to generate responsive imagery in real time

Each element was tested and refined through rehearsal. For example, when lyrics described soft hands or celestial beauty, we generated images like blooming lotus petals or ethereal light fields. These visuals were not literal translations. They were emotional cues that supported the tone and metaphor of the song.

Iteration Through Testing

We conducted small group test runs with peers and instructors to evaluate clarity and emotional impact. Key feedback helped us:

• Adjust the brightness and animation pacing of visuals

• Refine prompts to maintain abstraction without losing meaning

• Recalibrate how long the pre-performance context should last

Through this iterative process, we developed a system that felt less like a projection and more like a co-performer. The AI visuals would follow the dancer, bloom with her gestures, and dissolve as the story shifted.

Deliver

Bringing the Story to the Stage and Measuring Cultural Impact

After weeks of ideation and technical refinement, our final exhibition brought all components together in a one-day gallery performance. The experience was designed to guide the audience through layers of understanding—from introduction to emotional immersion.

Final Exhibition Design

The performance space was set up to support cultural framing from the moment guests walked in:

• TV slide loop introduced the dance form, narrative themes, and central characters

• Projection-mapped altar animated and labeled sacred objects to create familiarity

• Pamphlets with lyrics and translations gave audiences a narrative guide they could follow before and after the performance

Once seated, guests watched a live Kuchipudi performance of Alamelumanga Hari Antaranga with generative AI visuals unfolding in real time. Each movement and lyric triggered imagery that deepened emotional resonance—without simplifying or interrupting the dance.

Audience Response and Evaluation

We conducted a post-performance survey using QR codes and received feedback from over twenty participants. Highlights included:

• 90 percent of respondents said the performance helped them better understand Kuchipudi and its cultural context

• Many praised the AI visuals for being beautiful, poetic, and emotionally aligned without being distracting

• Multiple people expressed interest in seeing more performances like this that integrate tradition with technology

Select feedback included:

“I love love loved the AI visuals following the dancer. I’ve never seen anything like it.”

“The altar was beautiful and intriguing. I’d love to interact with it more.”

Takeaways and Reflection

This project reinforced that tradition and innovation are not opposites. With care and collaboration, technology can support cultural integrity. The AI system we developed was not just reactive—it was expressive. It listened, responded, and respected the space.

Key lessons I’m carrying forward:

• Design with empathy means understanding emotional barriers, not just informational ones

• Abstraction is a powerful tool when literal explanation risks flattening cultural nuance

• Collaborative design builds trust between disciplines and traditions

Expressions of Alamelu is not just a performance—it’s a prototype for how immersive technology can create new bridges into sacred traditions.