Work
DIANA CHALAKOVA
About
Contextpath
I was invited to participate in a one-day hackathon held by Anthropic and The Metropolitan Museum of Art at the Met, coinciding with the release of Claude Opus 4.8. The project tackled a simple but meaningful problem: the Asian Arts collection at the Met contains some of the oldest artifacts in the building, yet much of Western society lacks the cultural frame of reference to fully appreciate them in a way they might the Greek or Roman galleries. Using Claude, I built an experience that delivers that depth of context to visitors without pulling them away from the piece in front of them. I designed it in Figma and used Claude Code to bring it to life for a live demo.
Role
Lead UX Designer
Timeline
1 day hackathon
Tools
Figma
Claude Code
Hackathon Prompt
How might we encourage museum visitors to engage deeper with the MET’s Asian Art collection?
The Metropolitan Museum of Art in New York City’s Asian Art collection contains over 34,000 digitized works of art from 4000 BCE to the present day. However, some of the work contains symbols, ideas, or techniques that are unfamiliar to contemporary or non-expert audiences, from calligraphic writing to fantastic beasts.
So, the MET invited me to explore ways AI can invite audiences to further engage with this collection, on-site or online. They described this as either creating something in a dialog with a single, detail-rich work, or being inspired by or using data from large swaths of the collection
Problem
The missing piece: context.
I went and interviewed four willing participants, two who were exploring the Asian Art collection already and two who were in a different part of the museum. I asked them respectively what drew them to the collection and why they had or hadn’t explored the collection yet and engaged in further open-ended and non-leading questions through our conversations. The following are quotes from those conversations concerning the Asian Art collection specifically.
SOLUTION
Using Claude integration to provide a depth of context to an artifact without losing face-to-face experience with the piece.
The Asian Arts collection at the museum dates back to 3200–2000 BCE with ritual objects (bi discs) from the Neolithic Liangzhu culture – needless to say, that is a very, very long time ago. Such an incredibly rich and long history exists in all of the artifacts in the collection, yet much of Western society does not understand the context or the history behind the pieces like they might the Greek or Roman collection. So, we decided to bridge that gap by providing that depth of context through Claude and Anthropic’s services.
The app also pulls from the MET’s database of other works that have similar iconography or time in history to encourage engagement with other parts of the museum and highlight that we are, and have always been, all connected despite our geographical place in the world. We see opportunity for a connection to contemporary works as well.
underlying concept
An accessory to face-to-face experience with artifacts.
Let’s say that a museum visitor who has ContextPath views this artifact of the Standing Lion from the Tang Dynasty and would like to learn more about it in the context of when it was created.
They would scan the item, Claude would pull from the museum’s database to identify it, and then the user could learn more about the key words used in the label in that historical context to build understanding of the piece itself.
This could go on for as long as it’s interesting to the user. Important to this experience though is that users cannot view the piece through their phone, nor look at past piece information viewed without being in front of the piece. This way, we do not replace the significance of face-to-face connection with artifacts in museums.
The integration of AI in art and museum spaces often receives public pushback, so although it would not be a secret that the scan and database are powered by Claude, this would be discreet and enable a rabbit hole into more knowledge for the greater good which is less likely to create controversy.
Finally...
Thank you to Team Anthropic and the MET!
I'd like to close by extending my heartfelt thanks to the two remarkable institutions that made this project possible.
To the team at Anthropic — thank you for building tools that genuinely expand what's creatively and intellectually possible, and for bringing them into spaces like this one. Beyond the technology, the conversations I got to have with people from the Anthropic team on the day were so thoughtful, generous, and genuinely inspiring. Your support throughout the hackathon, and your belief that AI can be a meaningful collaborator in the arts, made all the difference.
To The Metropolitan Museum of Art — thank you for opening your doors, your collection, and your curiosity to us. There is something quietly profound about being invited to think about the future inside one of the world's great temples to human creativity.
Work
DIANA CHALAKOVA
About
Contextpath
I was invited to participate in a one-day hackathon held by Anthropic and The Metropolitan Museum of Art at the Met, coinciding with the release of Claude Opus 4.8. The project tackled a simple but meaningful problem: the Asian Arts collection at the Met contains some of the oldest artifacts in the building, yet much of Western society lacks the cultural frame of reference to fully appreciate them in a way they might the Greek or Roman galleries. Using Claude, I built an experience that delivers that depth of context to visitors without pulling them away from the piece in front of them. I designed it in Figma and used Claude Code to bring it to life for a live demo.
Role
Lead UX Designer
Timeline
1 day hackathon
Tools
Figma
Claude Code
Hackathon Prompt
How might we encourage museum visitors to engage deeper with the MET’s Asian Art collection?
The Metropolitan Museum of Art in New York City’s Asian Art collection contains over 34,000 digitized works of art from 4000 BCE to the present day. However, some of the work contains symbols, ideas, or techniques that are unfamiliar to contemporary or non-expert audiences, from calligraphic writing to fantastic beasts.
So, the MET invited me to explore ways AI can invite audiences to further engage with this collection, on-site or online. They described this as either creating something in a dialog with a single, detail-rich work, or being inspired by or using data from large swaths of the collection
Problem
The missing piece: context.
I went and interviewed four willing participants, two who were exploring the Asian Art collection already and two who were in a different part of the museum. I asked them respectively what drew them to the collection and why they had or hadn’t explored the collection yet and engaged in further open-ended and non-leading questions through our conversations. The following are quotes from those conversations concerning the Asian Art collection specifically.
SOLUTION
Using Claude integration to provide a depth of context to an artifact without losing face-to-face experience with the piece.
The Asian Arts collection at the museum dates back to 3200–2000 BCE with ritual objects (bi discs) from the Neolithic Liangzhu culture – needless to say, that is a very, very long time ago. Such an incredibly rich and long history exists in all of the artifacts in the collection, yet much of Western society does not understand the context or the history behind the pieces like they might the Greek or Roman collection. So, we decided to bridge that gap by providing that depth of context through Claude and Anthropic’s services.
The app also pulls from the MET’s database of other works that have similar iconography or time in history to encourage engagement with other parts of the museum and highlight that we are, and have always been, all connected despite our geographical place in the world. We see opportunity for a connection to contemporary works as well.
underlying concept
An accessory to face-to-face experience with artifacts.
Let’s say that a museum visitor who has ContextPath views this artifact of the Standing Lion from the Tang Dynasty and would like to learn more about it in the context of when it was created.
They would scan the item, Claude would pull from the museum’s database to identify it, and then the user could learn more about the key words used in the label in that historical context to build understanding of the piece itself.
This could go on for as long as it’s interesting to the user. Important to this experience though is that users cannot view the piece through their phone, nor look at past piece information viewed without being in front of the piece. This way, we do not replace the significance of face-to-face connection with artifacts in museums.
The integration of AI in art and museum spaces often receives public pushback, so although it would not be a secret that the scan and database are powered by Claude, this would be discreet and enable a rabbit hole into more knowledge for the greater good which is less likely to create controversy.
Finally...
Thank you to Team Anthropic and the MET!
I'd like to close by extending my heartfelt thanks to the two remarkable institutions that made this project possible.
To the team at Anthropic — thank you for building tools that genuinely expand what's creatively and intellectually possible, and for bringing them into spaces like this one. Beyond the technology, the conversations I got to have with people from the Anthropic team on the day were so thoughtful, generous, and genuinely inspiring. Your support throughout the hackathon, and your belief that AI can be a meaningful collaborator in the arts, made all the difference.
To The Metropolitan Museum of Art — thank you for opening your doors, your collection, and your curiosity to us. There is something quietly profound about being invited to think about the future inside one of the world's great temples to human creativity.
Work
DIANA CHALAKOVA
About
Contextpath
I was invited to participate in a one-day hackathon held by Anthropic and The Metropolitan Museum of Art at the Met, coinciding with the release of Claude Opus 4.8. The project tackled a simple but meaningful problem: the Asian Arts collection at the Met contains some of the oldest artifacts in the building, yet much of Western society lacks the cultural frame of reference to fully appreciate them in a way they might the Greek or Roman galleries. Using Claude, I built an experience that delivers that depth of context to visitors without pulling them away from the piece in front of them. I designed it in Figma and used Claude Code to bring it to life for a live demo.
Role
Lead UX Designer
Timeline
1 day hackathon
Tools
Figma
Claude Code
Hackathon Prompt
How might we encourage museum visitors to engage deeper with the MET’s Asian Art collection?
The Metropolitan Museum of Art in New York City’s Asian Art collection contains over 34,000 digitized works of art from 4000 BCE to the present day. However, some of the work contains symbols, ideas, or techniques that are unfamiliar to contemporary or non-expert audiences, from calligraphic writing to fantastic beasts.
So, the MET invited me to explore ways AI can invite audiences to further engage with this collection, on-site or online. They described this as either creating something in a dialog with a single, detail-rich work, or being inspired by or using data from large swaths of the collection
Problem
The missing piece: context.
I went and interviewed four willing participants, two who were exploring the Asian Art collection already and two who were in a different part of the museum. I asked them respectively what drew them to the collection and why they had or hadn’t explored the collection yet and engaged in further open-ended and non-leading questions through our conversations. The following are quotes from those conversations concerning the Asian Art collection specifically.
Solution
Using Claude integration to provide a depth of context to an artifact without losing face-to-face experience with the piece.
The Asian Arts collection at the museum dates back to 3200–2000 BCE with ritual objects (bi discs) from the Neolithic Liangzhu culture – needless to say, that is a very, very long time ago. Such an incredibly rich and long history exists in all of the artifacts in the collection, yet much of Western society does not understand the context or the history behind the pieces like they might the Greek or Roman collection. So, we decided to bridge that gap by providing that depth of context through Claude and Anthropic’s services.
The app also pulls from the MET’s database of other works that have similar iconography or time in history to encourage engagement with other parts of the museum and highlight that we are, and have always been, all connected despite our geographical place in the world. We see opportunity for a connection to contemporary works as well.
underlying concept
An accessory to face-to-face experience with artifacts.
Let’s say that a museum visitor who has ContextPath views this artifact of the Standing Lion from the Tang Dynasty and would like to learn more about it in the context of when it was created.
They would scan the item, Claude would pull from the museum’s database to identify it, and then the user could learn more about the key words used in the label in that historical context to build understanding of the piece itself.
This could go on for as long as it’s interesting to the user. Important to this experience though is that users cannot view the piece through their phone, nor look at past piece information viewed without being in front of the piece. This way, we do not replace the significance of face-to-face connection with artifacts in museums.
The integration of AI in art and museum spaces often receives public pushback, so although it would not be a secret that the scan and database are powered by Claude, this would be discreet and enable a rabbit hole into more knowledge for the greater good which is less likely to create controversy,
Finally...
Thank you to Team Anthropic and the MET!
I'd like to close by extending my heartfelt thanks to the two remarkable institutions that made this project possible.
To the team at Anthropic — thank you for building tools that genuinely expand what's creatively and intellectually possible, and for bringing them into spaces like this one. Beyond the technology, the conversations I got to have with people from the Anthropic team on the day were so thoughtful, generous, and genuinely inspiring. Your support throughout the hackathon, and your belief that AI can be a meaningful collaborator in the arts, made all the difference.
To The Metropolitan Museum of Art — thank you for opening your doors, your collection, and your curiosity to us. There is something quietly profound about being invited to think about the future inside one of the world's great temples to human creativity.