Georgia Tech’s faculty startup engine Quadrant-i, together with the Space Research Institute (SRI), launched the first cohort of the CreationsVC Space Fellows Program. Funded by space technology venture capital firm CreationsVC, the program enables faculty to explore promising early-stage innovations and their potential for future commercial impact. 

“This first set of CreationsVC Fellows offers an exciting cross-section of innovative hardware and software technologies built on Georgia Tech’s legacy of space exploration, hardware development, and product commercialization,” said Jud Ready, SRI executive director. 

In the first year of the three-year program, CreationsVC provides $125,000 to promote and accelerate innovations that have both space and terrestrial applications. The series offers participants training focused on customer discovery, engaging and compelling storytelling, value proposition design and quantification, and lean/agile project/product management.

“CreationsVC is centered on a deep appreciation for innovation and big thinking,” said Steve Braverman, co-founder and managing partner of CreationsVC. “We felt this was the right time to align our efforts in sourcing and supporting dual-value technologies that will have an impact on both Earth and space.” 

The six startups tackle real-world space research problems like supply chain management, how artificial intelligence works in space, and navigation.

“We are excited CreationsVC is providing us with an opportunity to try new approaches to accelerate deep tech development,” said Jonathan Goldman, Quadrant-i’s director. “These are the toughest kinds of startups to build, and we look forward to the learning we will gain from forcing our innovators out of their comfort zones to embrace some new and valuable skills.”

Meet the cohort:
 

Company: CIMTech.ai
 

Founders: Shimeng Yu, James Read

School: School of Electrical and Computer Engineering (ECE)

Objective: To develop energy-efficient, radiation-tolerant artificial intelligence processors using a persistent type of ferroelectric memory. The startup aims to improve applications requiring high power efficiency, such as battery-powered devices and space-based systems.

Why Q-i: “The advantage of Q-i is in helping technical founders turn their research into products that solve customers’ problems,” noted James Read. “For us, that means talking with potential customers and hearing their pain points directly from the source. Now we’re use that information to build a convincing narrative around our startup’s value for stakeholders and investors.” 

Company: SkyCT
 

Founders: Morris Cohen, Matthew Strong

School: ECE

Objective: To provide up-to-date mapping of the electrical properties of the upper atmosphere, with applications to GPS-free navigation, long-range communication, and satellite and launch vehicle viability. The startup uses the radio energy released by lightning strikes to create this map. 

Why Q-i: “This weird region about 50 miles up from Earth’s surface is both really hard to track and measure, and also impacts a surprising array of applications,” said Cohen. “It’s sometimes called the `ignorosphere’ because of how difficult it is to measure, and it’s time we change that.” 

Company: Penumbra Autonomy
 

Founders: Panagiotis Tsiotras, Juan Diego Florez-Castillo, Iason Velentzas 

School: Daniel Guggenheim School of Aerospace Engineering (AE)

Objective: To commercialize algorithms that help spacecraft maneuver when they have limited information on their environment. The algorithms use state-of-the-art computer vision and localization techniques. This could benefit manufacturing, assembly, and refueling in orbit, as well as enable monitoring, situational awareness, and debris removal. 

Why Q-i: “The program offers a conduit to entrepreneurship opportunities and spinoff companies in the space domain by providing guidance and commercialization ‘know-how,’” said Panagiotis Tsiotras. 

Company: TerraMorph

 
Founders: Yashwanth Kumar Nakka, Sadhana Kumar, Vincent Griffo, Sachin Kelkar

School: AE

Objective: To create an autonomous rover platform with adaptive, reconfigurable mobility. The rover will implement software and sensing algorithms to automatically detect terrain type and improve traction and energy usage. This could be used on the moon or Mars, or even terrestrial search and rescue. 

Why Q-i: “TerraMorph was developed to address fundamental challenges in mobility and autonomy across uncertain terrain,  but successfully translating that work into impact requires creative guidance, critical feedback, and experienced perspectives beyond the lab,” said Yashwanth Kumar Nakka. “Q-i’s culture of leading by example and fostering strong, ethical teams aligns closely with how we want to build TerraMorph: iteratively, thoughtfully, and with a focus on real-world deployment.” 

Company: OpenWerks
 

Founders:  Shreyes Melkote, Mike Yan

School: George W. Woodruff School of Mechanical Engineering

Objective: To deliver real-time manufacturing supply chain visibility for the space and national security industries. OpenWerks technology aims to dramatically reduce current sourcing cycles from eight months down to weeks by connecting corporate buyers directly with verified supplier manufacturing capability and capacity data. 

Why Q-i: “From the very beginning, principals at VentureLab and  Q-i offered a clear pathway to translate academic research into a viable business,” said Mike Yan. “Their reputation for guiding Georgia Tech startups through both business and technology derisking, combined with their comprehensive ecosystem of programs and coaches, made them the natural partner for our entrepreneurial journey.”

Company: 8Seven8
 

Founders: Chandra Raman

School: School of Physics

Objective: To manufacture quantum hardware in Georgia. 8Seven8 aims to put high-precision atomic clocks and gyroscopes on a chip for applications ranging from aircraft navigation to industrial automation.  

Why Q-i: “They have mentored me and my students through the commercialization process, providing opportunities such as the Space Fellows Cohort,” Chandra Raman said. “One of my former students, Alexandra Crawford, gained valuable business experience through a Q-i entrepreneur’s assistantship, and is now working at 8Seven8 full-time. They have also guided me through the process of obtaining funding through the Georgia Research Alliance for our commercialization effort.”

Five faculty members from Georgia Tech have been elected as senior members of the National Academy of Inventors (NAI). As members, they are recognized as distinguished academic inventors with a strong record of patenting technologies, licensing IP, and commercializing their research. Their innovations have made, or have the potential to make, meaningful impacts on society. 

 “The election of our faculty members to this prestigious association is a powerful affirmation of the innovative research happening at Georgia Tech,” said Raghupathy “Siva” Sivakumar, chief commercialization officer at Georgia Tech. “Their work to take research to market reflects the growing importance of invention in addressing society’s most complex challenges. This recognition signals the strength of the commercialization ecosystem at Georgia Tech to advance impactful research, encourage innovation, and prepare the next generation of inventors.” 

The 2026 Georgia Tech NAI senior members are: 

• Jason David Azoulay, associate professor, School of Materials Science and Engineering School and School of Chemistry and Biochemistry
• Jaydev Prataprai Desai, professor and cardiovascular biomedical engineering distinguished chair, Wallace H. Coulter Department of Biomedical Engineering
• David Frost, Elizabeth and Bill Higginbotham Professor and Regents’ Entrepreneur, School of Civil and Environmental Engineering
• Chandra Raman, Dunn Family Professor of Physics, School of Physics
• Aaron Young, associate professor, George W. Woodruff School of Mechanical Engineering

Jason David Azoulay

Azoulay is recognized for pioneering new classes of functional materials through innovative polymer synthesis, heterocycle chemistry, and polymerization reactions. His work spans electronic, photonic, and quantum materials, device fabrication, and chemical sensing for environmental monitoring. He has demonstrated new classes of organic semiconductors with infrared functionality and holds nine issued U.S. patents. Azoulay is the Georgia Research Alliance Vasser-Woolley Distinguished Investigator and holds a joint appointment in the School of Chemistry and Biochemistry. 

Jaydev Prataprai Desai

Desai is recognized for advancing medical robotics and translational biomedical innovation with inventions spanning robotically steerable guidewires for endovascular interventions, minimally invasive surgical tools, MEMS sensors for cancer diagnosis, and rehabilitation robotics for people with motor impairments. He is the founding editor-in-chief of the Journal of Medical Robotics Research, has authored more than 225 peer-reviewed publications, and serves as the Director of Georgia Center for Medical Robotics at Georgia Tech. Desai holds 15 U.S. and International patents.  

David Frost

Frost has built a career at the intersection of civil engineering research and entrepreneurship. A leader in the study of natural and human-made disasters and their impacts on infrastructure, he has founded two Georgia Tech-based software companies: Dataforensics, which offers tools for subsurface data collection and infrastructure project management, and Filio, an AI-powered mobile platform that supports visual asset management in construction and post-disaster reconnaissance. In 2023, Frost was named a Regents’ Entrepreneur by the University System of Georgia’s Board of Regents, a designation reserved for tenured faculty who have successfully taken their research into a commercial setting. He holds four U.S. patents.  

Chandra Raman

Raman is a physicist, inventor, and technology entrepreneur whose research on ultracold atoms is enabling a new generation of ultraprecise quantum sensing devices. He is the co-inventor of chip-scale atomic beam technology — a breakthrough that makes it possible to miniaturize quantum sensors for navigation and timing applications in environments where GPS fails, with uses spanning autonomous vehicles, aerospace, and national security. Raman holds six U.S. patents, three of which have been issued and two licensed. To bring his inventions to market, he founded 8Seven8 Inc., Georgia’s first quantum hardware company. He is a fellow of the American Physical Society and an advisor to national and space-based quantum initiatives. 

Aaron Young

Young directs the Exoskeleton and Prosthetic Intelligent Controls Lab, where he develops robotic exoskeletons and intelligent control systems to improve walking function and physical capability for people with mobility impairments and industrial safety applications. His research has been supported by major federal grants from the National Institutes of Health, and he holds three U.S. patents. Young works with Georgia Tech’s Office of Technology Licensing and Quadrant-i to advance promising technologies toward real-world adoption. 

About Georgia Tech’s Office of Commercialization 

The Office of Commercialization is the nexus of research commercialization and entrepreneurship at Georgia Tech, bringing leading-edge research and innovation to market. It comprises six key units — ATDC, CREATE-X, VentureLab, Quadrant-i, Technology Licensing, and Velocity Startups — that empower students and faculty to launch startups, manage intellectual property, and transform research ideas into positive societal impact. Learn more at commercialization.gatech.edu. 

About the National Academy of Inventors 

The National Academy of Inventors is a member organization comprising U.S. and international universities, and governmental and nonprofit research institutes, with over 4,000 individual inventor members and fellows spanning more than 250 institutions worldwide. It was founded in 2010 to recognize and encourage inventors with patents issued from the U.S. Patent and Trademark Office, enhance the visibility of academic technology and innovation, and translate the inventions of its members to benefit society. Learn more at academyofinventors.org. 

Abstract 

“It was a hypothesis. I was the experiment, and the hypothesis was proven true.” 

Can an inner-city student who grew up below the poverty line earn a Ph.D. and make a career in research? In theory, yes.  

The barriers are many. But literature suggests that early exposure to STEM and research opportunities can increase the odds for students in need.  

For Kendreze Holland, the idea of making it to college and earning an advanced degree was a hypothesis. Sure, theoretically it could be done — but in his own home, not everyone had even made it past high school.  

Often, the first question on the way to scientific discovery is: What if? What if a student like Holland received the right help at the right time? What if he was guided along the way by mentors who were leaders in their fields? What if he was given the opportunity to develop professional skills and make valuable connections? 

Holland asked himself: What if he could be the one to prove the hypothesis true? 

Introduction 

Holland grew up in northwest Atlanta, one of seven children raised by a single mother. Being one of so many children, most would struggle to stand out. But Holland always sought to be different.  

“My perpetual intention was to be less of a burden to my mother,” he said. “Since my mother’s education limited her abilities to help with my schoolwork, I went above the call of duty to stand out in academics.” 

His mother’s education was cut short in ninth grade so she could raise her first child, Holland’s older sister, and no one in his family had gone to college. In his mind, he had three career paths to choose from: football, hip hop, or retail.  

“Standing at a solid 5 foot 8, the first would have been difficult,” he joked. “And the latter two were not my calling.” 

Just like his mother, the course of his life changed in his ninth-grade year. For Holland, it began an academic journey he never expected.  

In 2012, he was attending B.E.S.T. Academy, an all-boys public school for grades six through 12 focused on business and STEM. Biology class was just another hour waiting to pass for the 15-year-old Holland, until the day two guest speakers from Georgia Tech walked into the room with “some weird apparatuses and mechanical chopsticks.” 

The two guests used the equipment — gel electrophoresis systems and pipettes — to show the boys what research can look like in real life. 

“This experience sparked within me a drive for science, and it was the first time I realized that I wanted to, and could, attain an advanced scientific degree,” Holland said.  

The two speakers were Manu Platt, a professor in the Wallace H. Coulter Department of Biomedical Engineering at Georgia Tech and Emory University, and Jerald Dumas, a postdoctoral researcher. Platt and Dumas were there to recruit students for a new program called Project ENGAGES within the Parker H. Petit Institute for Bioengineering and Bioscience (IBB).  

The program was co-founded by Platt and the late Robert M. Nerem, IBB’s founding executive director, to give students like Holland an opportunity to participate in real research projects that would hopefully plant a seed in the next generation of scientists.  

Students come from one of eight partner schools in Atlanta. Once accepted, they are connected to a Georgia Tech graduate student who mentors them and supervises their work, and they get paid to work in their assigned lab for one year.  

Project ENGAGES does more than expose students to STEM concepts and ideas. It equips them with the skills and knowledge to carry out their own independent research projects. They also have opportunities to establish connections with university faculty and industry representatives who can provide career guidance and support. 

Methods 

Though Holland didn’t meet the program’s age requirement in 2012, he applied again the next year and was accepted. During his junior and senior years of high school, he worked in Platt’s lab, where he aided with projects involving proteins, cell cultures, and antibodies.  

“Over the course of those two years, the growth I saw scientifically, professionally, and in maturity, all corroborated my belief that Kendreze was going far, and able to push past whatever goals and obstacles he comes up against,” said Platt, now the director of the Center for Biomedical Engineering Technology Acceleration housed within the National Institute of Biomedical Imaging and Bioengineering.  

Holland's experience sparked a love for science and a career-long connection with Georgia Tech. After high school, he graduated summa cum laude with a degree in chemistry from Georgia State University. As an undergraduate, he stayed connected with Tech and with IBB as a Petit Scholar, a yearlong mentorship program and research experience for top students around Atlanta. 

“I really wanted to stay close to home, and I felt like everything was in my backyard,” he said. “There are many people who come here from other places to Tech because of the great science that is going on. There’s something special about Atlanta, and I’m just getting the best of what I can from it.” 

He credits his time in Project ENGAGES with giving him the confidence and resilience to continue toward his goals. Like many others in the program, he was a first-generation college student with little to no guidance for his academic career. The holistic approach of Project ENGAGES provided professional development opportunities and standardized test preparation to ready him for life in college and beyond. 

“I knew I wanted to go to grad school, but I didn’t know I was going to do all these things,” he said. “Having that one goal sprouted a lot of side quests that just grew into something bigger.” 

After graduating from Georgia State in 2020, Holland was accepted into Georgia Tech’s Bioengineering Graduate Program as a doctoral student. In December 2025, he became the first Project ENGAGES alumnus to successfully defend his dissertation, and he is expected to graduate this spring. 

Lakeita Servance, assistant director of Outreach Initiatives at IBB, was the program manager for Project ENGAGES when Holland was accepted and cheered him on more than 10 years later as he presented his doctoral research. 

“As I sat in that room while he was defending his dissertation and sharing his research with all of us, I still reflected on that boy I saw at 16 years old,” she said. “It was this full circle moment to see him make it all the way back here. The investment we made over a decade ago has paid off in such a large way.” 

Results 

In addition to being the first in his family to go to college and earn an advanced degree, Holland received financial support from the National Science Foundation’s Graduate Research Fellowship Program; was awarded multiple prestigious fellowships, including FORD, GEM, and Herbert P. Haley; landed an internship with 3M Corporate Research Materials Laboratory; and served as a mentor in the Nakatani Research and International Experience for Students. He has published papers, led panel discussions, applied for patents, and presented his research at national conferences.   

“All that stemmed from Project ENGAGES,” he said. “And more importantly, I applied to be a mentor for the ENGAGES program.” 

Holland said some of his most meaningful experiences have come from being able to give back. He has served as a mentor, both formally and informally, to more than half a dozen students, some who come from backgrounds much like his own. 

“I wanted to give back to the program because it poured so much into me. They were able to get me all the way to the Ph.D. level, so I knew that I could use my grind to help other students.” 

Conclusion 

Having proved the hypothesis true, Holland is turning his focus to the future, considering his options in academia and corporate research while he continues to work as a postdoc at Georgia Tech.  

His research in John Blazeck’s lab focuses on cellular engineering using CRISPR gene editing technology to regulate gene profiles, meaning he and other researchers can turn certain genes up and others down to affect the way cells respond. Though he is currently working with yeast cells, he hopes that his research will translate into mammalian cells that could have more clinical applications.  

“In terms of diseases and disorders, you can use it to tune genes to help someone experiencing cancer by helping immune cells or stopping cancer cells from dividing rapidly,” he said. “You can also help other cells to survive longer, and longer cell viability means potentially a patient can survive longer.” 

What began as a presentation in a high school science class has led Holland to a future he never expected. Tequila Harris, professor in the George W. Woodruff School of Mechanical Engineering and co-director of Project ENGAGES, said his story shows others that they can do the same.  

“I believe his achievements will inspire and motivate generations of students to pursue dreams that they may not have known they had. Kendreze Holland has fundamentally shown others that there are multiple pathways to engage in STEM and that opportunities and access to advanced degrees can be attained by those willing to do the work.” 

Holland's story is symbolic of the ultimate goal for Project ENGAGES: to change the lives of talented young people who may never have had the opportunity to succeed.  

“That’s why I was so adamant about getting my Ph.D.,” he said, “to show that one could potentially overcome what they were going through to do something extraordinary.” 

 

Project ENGAGES is possible thanks to philanthropic support from our generous community: Donate here.

Written by: Anne Wainscott-Sargent

As artificial intelligence (AI) drives explosive growth in data centers, communities across the U.S. are facing rising electricity costs, new industrial development, and mounting strain on an aging power grid.

At Georgia Tech, several faculty members are approaching these sustainability challenges from different but complementary angles: examining how data center policy affects local communities, modeling how AI-driven demand reshapes regional energy systems, and building tools that help the public understand the tradeoffs embedded in grid planning. Together, their work highlights how better data, thoughtful policy, and public engagement can guide more resilient and equitable decisions in an AI-powered future.

AI’s Hidden Footprint: How Data Centers Reshape Communities

Ahmed Saeed studies the infrastructure most people never see. An assistant professor in the School of Computer Science and a Brook Byers Institute for Sustainable Systems (BBISS) Faculty Fellow, Saeed focuses on how data centers — the backbone of modern AI — are built, operated, and regulated, and what their growth means for host communities.

“Data centers are the infrastructure for our digital life, so more of them are necessary to keep doing what we’re doing,” he said.

Data center energy consumption could double or triple by 2028, accounting for up to 12% of U.S. electricity use, according to a report by Lawrence Berkeley National Laboratory. U.S. spending on data center construction jumped nearly 70% between May 2023 and May 2024, according to the American Edge Project.

Georgia is an AI data center hub, ranked fourth globally, with $4.6 billion in AI-related venture capital invested across 368 deals, the American Edge Project reported. At a recent town hall in DeKalb County, Georgia, Saeed helped residents connect AI’s promise to its local consequences. Training large AI models can require tens of thousands of graphics processing units (GPUs) running for days or weeks, driving an unprecedented wave of data center construction. AI-focused chips, he noted, can consume 10 to 14 times more power than traditional processors.

That demand often shows up as pressure on local infrastructure. Communities are increasingly concerned about electricity and water use, grid upgrades, and who ultimately pays. In Virginia, Saeed pointed to a legal dispute in which consumer advocates warned that data centers could raise electricity bills by 5% in the short term and up to 50% over time, while utilities argued those investments were inevitable and could benefit customers in the long run.

Environmental concerns add another layer. Saeed cited controversies over water use and backup diesel generators in states, including Georgia and Tennessee, alongside a recent Environmental Protection Agency (EPA) ruling that tightened generator regulations. While diesel generators are clearly harmful, he cautioned that long-term, rigorous evidence linking data centers to regional health impacts remains limited.

Saeed’s research aims to reduce those impacts directly. By optimizing how workloads are scheduled across large server fleets, his team has demonstrated power savings of 4 – 12%, a meaningful gain if U.S. data centers approach projected levels of up to 12% of national electricity use by 2028.

For Saeed, data centers are akin to highways: essential to modern life, disruptive to nearby communities, and shaped by policy choices. The question, he argues, is not whether AI infrastructure should exist, but how transparently and fairly it is built.

Economist Probes the Energy Costs of the AI Boom

While headlines often frame AI as an energy crisis, Georgia Tech environmental and energy economist and BBISS Faculty Fellow Tony Harding is focused on measuring its real — and uneven — impacts. Harding, an assistant professor in the Jimmy and Rosalynn Carter School of Public Policy, uses economic modeling to examine how AI adoption affects energy use, emissions, and local communities.

In recent work published in Environmental Research Letters, Harding and his co-author analyzed how productivity gains from AI could influence national energy demand. Their findings suggest that, at a macro level, AI-related activity may increase annual U.S. energy use by about 0.03% and CO₂ emissions by roughly 0.02%.

“Those numbers are small in the context of the overall economy,” Harding said. “But the impacts are highly uneven.”

That unevenness is evident in where data centers are built. While Northern Virginia remains the country’s top data center hub, with 343 operational data centers, states like Georgia, which currently has 94 operational data centers, are rapidly attracting facilities due to reliable power and favorable tax policies. 

Harding’s latest research focuses on local effects, asking why data centers cluster in urban areas, how they influence housing markets, what happens to electricity prices, and whether they exacerbate water stress. Early evidence suggests large facilities can increase local electricity rates, contributing to public backlash and regulatory response. In Georgia, the Public Service Commission has begun requiring new, high power draw customers (like data centers) to cover more of the costs associated with grid expansion.

Harding’s goal is to give policymakers better evidence to design incentives and guardrails. “To manage these technologies responsibly,” he said, “we need a clear picture of their intended and unintended consequences.”

Gamifying a Strained and Aging Power Grid

Daniel Molzahn is tackling another side of the problem: how to modernize an aging power grid under growing demand. Electricity demand is expected to rise about 25% by 2030, driven by data centers, electric vehicles, and broadscale electrification. At the same time, much of the U.S. electricity grid is nearing the end of its lifespan, with many transformers being decades old.

To make these challenges tangible, Molzahn, an associate professor in the School of Electrical and Computer Engineering, developed a browser-based game with a group of students through Georgia Tech’s Vertically Integrated Projects program called Current Crisis. Players take on the role of a utility decision-maker, balancing reliability, wildfire risk, renewable integration, and affordability.

The game grew out of Molzahn’s National Science Foundation CAREER award and reflects his belief that complex systems are best understood experientially. Its initial focus is wildfire resilience, modeling how grid infrastructure can both spark and suffer damage from fires.

But resilience comes at a cost. Burying power lines, for example, reduces wildfire risk but dramatically increases expenses. Players must confront the same tradeoffs utilities face: improve reliability or keep rates low.

Molzahn hopes the game will help students and the public grapple with the realities of planning future power systems. “These choices aren’t abstract,” he said. “They shape affordability, resilience, and our path toward a cleaner grid.”

The project now involves nearly 40 students from across campus, supported by Sustainability NEXT funding and a collaboration with Jessica Roberts, former BBISS Faculty Fellow and director of the Technology-Integrated Learning Environments (TILES) Lab in the School of Interactive Computing.

“As a learning scientist, I look at how to engage people with science and scientific data and get people having conversations they might not otherwise have,” says Roberts, who hopes the seed grant helps the team determine first that they are going in the right direction and, second, how to broaden the impact.

One student, Stella Quinto Lima, a graduate research assistant in Human-Centered Computing, has made the game the focus of her doctoral thesis. Through the game, she wants players to notice their misconceptions about the power grid, energy use, and AI, and to use critical thinking to identify, question, and possibly undo those misconceptions.

 “I hope that we can really engage adults and help them see it’s not black and white. The game is not only about power grids, but how AI affects the grid, how it affects our lives, and how it will impact our future.”

The team plans to expand the game’s features, use it in outreach programs, and analyze player decisions as a source of data to study energy-system decision-making.

“We want to change the conversation about power and power grid stability, reliability, and sustainability, Roberts said, “and find a way to get this message to a larger public.”

Geosynthetics are a category of materials—textiles, grids, membranes, composites, and more—that are used in infrastructure projects like roads, retaining walls and landfills.

Civil Engineering Professor David Frost, the new president of the Geosynthetic Institute, said geosynthetic materials are an important technology for engineers working to design more resilient infrastructure to withstand the increasingly severe natural disasters of the future.

“Geosynthetics are a resilience maker,” Frost said. “Whether to enhance the strength, alter the hydraulic conductivity, limit the deformation or control various rate processes, geosynthetics inherently augment the engineering properties of natural geomaterials.”

Read the rest of the article on the School of Civil and Environmental Engineering website. 

A Georgia Tech Ph.D. candidate is getting a boost to his research into developing more efficient multi-tasking artificial intelligence (AI) models without fine-tuning.

Georgia Stoica is one of 38 Ph.D. students worldwide researching machine learning who were named a 2025 Google Ph.D. Fellow.

Stoica is designing AI training methods that bypass fine-tuning, which is the process of adapting a large pre-trained model to perform new tasks. Fine-tuning is one of the most common ways engineers update large-language models like ChatGPT, Gemini, and Claude to add new capabilities. 

If an AI company wants to give a model a new capability, it could create a new model from scratch for that specific purpose. However, if the model already has relevant training and knowledge of the new task, fine-tuning is cheaper.

Stoica argues that fine-tuning still uses large amounts of data, and that other methods can help models learn more effectively and efficiently.

“Full fine-tuning yields strong performance, but it can be costly, and it risks catastrophic forgetting,” Stoica said. “My research asks if we can extend a model’s capabilities by imbuing it with the expertise of others, without fine-tuning?

“Reducing cost and improving efficiency is more important than ever. We have so many publicly available models that have been trained to solve a variety of tasks. It’s redundant to train a new model from scratch. It’s much more efficient to leverage the information that already exists to get a model up to speed.”

Stoica said the solution is a cost-effective method called model merging. This method combines two or more AI models into a single model, improving performance without fine-tuning.

On a basic level, Stoica said an example would be combining a model that is efficient at classifying cats with one that works well at dogs.

“Merging is cheap because you just take the parameters, the weights of your existing models, and combine them,” he said. “You could take the average of the weights to create a new model, but that sometimes doesn’t work. My work has aimed to rearrange the weights so they can communicate easily with each other.”

Through his Google fellowship, Stoica seeks to apply model merging to create a cutting-edge vision encoder. A vision encoder converts image or video data into numerical representations that computers can understand. This enables tasks such as image or facial recognition and generative image captioning.

“I want to be at the frontier of the field, and Google is clearly part of that,” Stoica said. “The vision encoder is very large-scale, and Google has the infrastructure to accommodate it.”