DALI Lab Drives Innovation

News subtitle

Technigala and Y Combinator events feature tech projects from Dartmouth.

Image
Image
Technigala graphic
Body

Current and former members of the Digital Applied Learning and Innovation Lab at Dartmouth have been showcasing their work at recent events.

Several Dartmouth students and DALI alums who founded companies were among the most recent cohort accepted into Y Combinator, a well-known startup accelerator based in Silicon Valley.

“DALI is a student-driven curiosity lab. We design and build technology solutions for a diverse array of problems for partners from around the world,” says DALI Director Tim Tregubov, a senior lecturer in the Department of Computer Science.

And among the 56 demos at the Fall ’24 Technigala, held on Nov. 19 at the Class of ’82 Engineering and Computer Science Center, was a project with Benchify—one of the five startups that was accepted into this summer’s three-month Y Combinator program.

The DALI lab hosts Technigala, a showcase of final computer science class projects and initiatives that students are working on with partners, at the end of every term. The event is sponsored by the computer science department, the Digital Arts Program, and DALI Lab as part of the Magnuson Center for Entrepreneurship.

Co-founded by Juan Castaño ’19 and Max von Hippel, Benchify automates software testing processes to write bug-free code, fast.

“The idea was to create a really easy-to-use software testing product. What makes us different is that we’re using techniques typically used in academia in aerospace and chip design,” says Castaño, who worked on DALI projects all four years of his Dartmouth career.

Image
Student trying virtual headset/eyewear
Attendees at Technigala try out virtual reality apps developed by DALI Lab participants. (Photo by Eli Burakian ’00)

Entry into the Y Combinator program is extremely competitive, with an average acceptance rate of less than 3%. Successful applicants receive $500,000 in funding and access to professional networks and mentorship to help the startup take off. Notable program alumni include Dropbox, DoorDash, AirBnB, and Reddit.

Along with Benchify, this year’s summer batch with DALI ties included: 

  • Lumenary, an AI-powered tool to accelerate financial research, founded by Vivek Hazari ’22, Daniel DiPietro ’22, Ziray Hao ’22, and John McCambridge ’23
  • Pumpkin, an AI web browser built by Sam Crombie ’23
  • CodeViz, an AI agent that creates a visual interface for code, founded by Liam Prevelige ’23 and Will McCall ’23
  • And deepsilicon, a company founded by computer science and engineering major Abhinav Reddy ’27 and Alexander Nanda ’27, who is studying computer science and physics. They are focused on creating simple software and hardware to run neural networks faster and cheaper.

Reddy, who worked as a software engineer for a year before starting at Dartmouth, says that DALI does a great job in replicating how teams operate in the corporate setting. “DALI sets up students for the company work environment, you basically know everything you need to know when you join. The onboarding process is very smooth,” he says.

Benchify is currently working with a DALI team to expand their tool and add new features for a more streamlined debugging process. “I loved working at DALI, I know how it works, and I was excited by the chance to partner with them,” says Castaño.

Learning by building

The DALI Lab recruits students to work in small teams as developers and designers to build digital solutions to complex, real-world problems. Designed as a startup-like experiential learning program, DALI was co-founded by Daniel Rockmore, the William H. Neukom 1964 Distinguished Professor of Computational Science, Lorie Loeb, research professor of computer science, and Tregubov.

Image
Rolando Coto Solano
Linguistics professor Rolando Coto Solano talks with an attendee at Technigala. Coto Solano uses machine language technology to build automatic speech recognition tools for Indigenous languages. (Photo by Eli Burakian ’00)

It’s common for students to work on one or more projects over many terms, even years, Tregubov says, making the experience different from a class project or capstone course. “By the time a DALI student graduates, they have designed, built, and shipped several projects/products of significant scope,” he says.

Recent examples include a virtual reality app called Veridium that brings chemistry lessons to life in the classroom and a smart microscope created in partnership with Dartmouth Health dermatopathologist Aravindhan Sriharan that allows clinicians to diagnose disorders from digital images.

Sriharan and the DALI team showcased the latest version of the SmartScope at the fall Technigala alongside newer projects such as Greta 2.0 and DeanHub.

Greta 2.0 was designed as a real world replica of the AI sibling in Matthew Libby’s play Sisters, performed at the Northern Stage in White River Junction in late September, that explored the relations between humans and technology.

Created by a team of students and staff members from the DALI Lab, the chatbot interacted with theatergoers offstage.

“The idea was to deepen people’s connection with the play. It was very interesting to see the questions that users asked and the quirky responses Greta 2.0 gave,” said the team’s product manager Richa Ranjan, Thayer ’25.

DeanHub, an application being developed for Dartmouth Information, Technology & Consulting, pulls together student course information for deans on an easy to visualize platform.

Projects using augmented reality and virtual reality developed as part of a computer science course were a major draw. Event visitors tried out a variety of VR apps—one created dynamic 3D art based on word prompts they chose while another let users explore a dark labyrinth that sprung jump scares and other spooky surprises as they moved around.

Transforming ideas into impact

Away from the crowded atrium at Technigala, a small group gathered in a conference room inside the DALI lab. Rolando Coto Solano, assistant professor of linguistics and adjunct assistant professor of computer science, met with the first cohort of students selected for the E.E. Just DALI Internship.

During the two-year program, students will gain hands-on experience working at the interface of technology and design on projects that will positively impact underrepresented and marginalized communities.

Coto Solano uses machine language technology to build automatic speech recognition tools for Indigenous languages such as Cook Islands Māori. These tools can help transcribe large volumes of video recordings that chronicle stories and cultures of these communities to create educational materials and help keep these languages alive.

“I need all the help I can get to develop better interface design for these tools and make them more effective,” says Coto Solano. Through this program, students will develop technical skills while working to build tools that are going to be impactful to people on the ground, he says.

This spirit of taking an idea, working it through, and finding impact is central to DALI’s mission, says Tregubov.

Evidently excited to see so many familiar names in the Y Combinator roster this year, their success underscores the experience they gain as undergraduates at Dartmouth and DALI, he says.

“Dartmouth Computer Science and DALI now have students who are starting up their own thing to try to change the world,” says Tregubov. “It’s awesome.”