Keti Sulamanidze

For Keti Sulamanidze, AI is more than a passing wave. The Bachelor in Computer Science and Artificial Intelligence student saw its potential for business and society, and dedicated herself to building the skills to make a difference through tech—all while inspiring others to do the same.

One of Keti’s biggest achievements at IE University is founding the IE Google Developer Student Club. Now serving as co-president, she and her entirely student-led team have managed to build a vibrant community of tech enthusiasts through a roster of exciting projects, events and sustainability-driven tech challenges. So when the opportunity to work for the tech giant came up at the tail-end of her third year, Keti knew immediately what she had to do.

“I worked as a software engineer intern on the Google Core Compute team,” she says. The prestigious summer internship took Keti to Warsaw, Poland, for 13 weeks, marking an excellent start to a computer science career. But besides cultural immersion, the internship allowed this student to contribute to developing a new, cutting-edge Google innovation.

Every year, Google allows student interns from all over the world to find practical solutions to a specific challenge. Keti’s project involved “investigating the feasibility of semantically clustering API responses. Besides the benefits of working alongside some of the world’s best engineers, it also gave her a chance to use the latest technologies hands-on, dipping her toes into pioneering applications of generative AI and machine learning.

Generating the right insights required utilizing various technologies. To start, Keti turned to Google’s Gemini AI chatbot to produce quality bulk responses. She then leveraged a BERT-based (Bidirectional Encoder Representations from Transformers) model to achieve refined, more efficient and domain-specific results. Later, she visualized her findings through advanced t-SNE (t-distributed stochastic neighbor embedding) methodologies, ensuring a better user experience.

However, engaging with potential users throughout the development stage was what truly took this project to the next level. “It was valuable in gathering feedback, which we used to refine the product even further.” The final steps involved validating each part of the pipeline through various metrics, as well as documenting the whole process for further advancements.

Keti credits the diverse responsibilities of her internship with improving both her technical and soft skills. The experience didn’t just solidify her grasp of machine learning models, it also allowed her to practice calibrating them using contrastive learning techniques. She also gained valuable practical skills in leveraging generative AI models like Google’s Gemini for high-quality API responses. Meanwhile, working with t-SNE showed her how to share technical information in user-friendly formats through hierarchical clustering and next-level data visualization. “It was a true research-like experience,” she shares.

On the other hand, Keti has noticed a sharp boost in her problem-solving skills since completing her Google internship. This is a direct benefit of “working on a complex project from scratch and finding innovative solutions along the way,” she admits. Indeed, building a new product comes with a unique set of challenges. For Keti, one of them was coming up with a completely new way to tackle the issue of semantically clustering API responses—an area that’s still barely explored.  

Despite these difficulties, the task opened up new opportunities for collaboration with leading AI professionals worldwide. “I consulted with colleagues and sought advice from machine learning experts, including some based in Zurich,” explains Keti. “This collaborative approach not only helped me overcome the challenge, but also enriched the project with diverse perspectives and solutions.” In addition, it further enhanced her project management toolkit, practically demonstrating the value of documentation and of integrating user feedback into the product development process.

Keti says that the Bachelor in Computer Science and Artificial Intelligence was instrumental in preparing her for such a high-level internship. “The skills I developed in my classes helped me navigate various aspects of the project, from building efficient solutions to understanding, fine-tuning and validating machine learning models.” She points out several courses in particular, including Software Development and DevOps, AI: Reasoning & Problem Solving, and AI: Chatbots & Recommendation Engines, among others. “These classes provided me with a strong foundation in both software engineering and AI.”

Admittedly, Google’s friendly, creative company culture made it easy for Keti to excel. “The team I worked with was exceptionally talented, incredibly friendly, creative and always willing to help, making my learning curve smoother and my overall experience richer.” 

Engaging with IE Talent & Careers before starting her internship equipped her even more, from career advice to mock interviews, information sessions and helpful career resources and guidance. The team also facilitated networking opportunities with IE Alumni already working at Google. “These connections were invaluable as they provided me with firsthand insights into the company culture, the specifics of the role and the application process itself.”

As Keti enters the last stage of her Bachelor in Computer Science and Artificial Intelligence, she’s determined to expand her skill set in order to lead in the industry’s dynamic future. She’s sure that IE University will continue empowering her with the tools to succeed.