Love it

Looking forward to making more music for this project!

This project has been a super fun learning experience but has also posed huge challenges. There has been countless technical challenges we have overcome but project has also been very intellectually strenuous as we make huge decisions and commitments for our future and the future of the project.

It's always such a pleasure to work with my talented collaborators @anushhkathakur.bsky.social and Mateo Larrea. We had a lot of fun implementing changes and dreaming about the future of the project together. We also filmed a vision video for a new direction the twins will go. Exciting progress ...

Spent a lot of time working on pitch decks to start presenting the human digital twins project to investors and getting feedback from mentors. This has been a very formative time for the project dictating the different markets we are tapping into and how the project will progress moving forward.

I added a feature that is a huge game changer in the conversational nature of the twin: auto greetings. Once the twin detects a face that is engaged the twin greets them automatically to start conversation. I have given the twin the sense of sight and now I am teaching it how to use this new input!

Optimized my computer vision system and added some great new features including an engagement state machine and multi person priority. This system dictates how natural the interaction feels and it just keeps getting smoother. Moving away from push to talk has made a huge difference.

So excited to announce @anushhkathakur.bsky.social, Mateo Larrea and I will be attending #CHI2026 is in Barcelona, Spain April 13 -17, where we will be bringing our human digital twins project as an interactive exhibition. We are super excited and grateful for the opportunity! @chi.acm.org

Devlog 5: After a lot of data smoothing and optimization, the CV was integrated into Unreal Engine to trigger different animation states based on the engagement of the viewer. I then filmed this interaction in a demo of all the cool features my collaborators and I added for our CHI submission!

Devlog 4: Continuing work on the CV interaction. I have overridden our old push to talk system and established a new one that uses webcam to detect engagement of the user. Its working but the data is too noisy. It requires a lot of smoothing, tweaking and optimization to implement into UE!

Devlog 3: I am working on the implementation of Computer vision as a way to streamline the interactions had with the twin. With this feature the twin will be able to recognize the user's face and determine their level of engagement. Based on this it will turn on the mic and listen for conversation.

Devlog 2: Slow and steady progress made on the Twins project in the second week of the academic quarter. We have delegated tasks in order to add new features we can demonstrate for our CHI submission. We are submitting to the "Interactive Demos" category and are excited to get back to work!

Devlog 1: After presenting the Augtwins at NEURIPS 2025 we returned to Savannah eager to make progress on our pitch deck and plan how we would approach our next conference submission. This time to CHI 2026. Looking forward to taking this project onto market one day at a time.