Utkarsh Trehan
Data Scientist @Amadeus
Signup · Get unlimited contacts
WORK HISTORY
Data Scientist @Amadeus
Nice, FR
Leading the end-to-end development of an internal Room Type Mapping (RTM) solution to standardize hotel room data across suppliers. Responsible for solution design, implementation strategy, and hands-on development; defined KPIs with the project manager, collaborated with data/component owners to align upstream inputs, and worked with architects on production deployment planning.• Developed a deep learning–based image classification model to assign hotel images to standardized Online Travel Agency (OTA) categories, significantly improving consistency over supplier-provided labels.• Migrated existing data engineering operations from on-premise servers to Azure Databricks, enabling more efficient and scalable data processing.• Mentored an intern in developing a proof of concept to enhance hotel descriptions using Large Language Models (LLMs), prompt engineering, and supporting NLP techniques such as entity extraction and text summarization, applied across structured and unstructured data.
EDUCATION
EURECOM
Master's degree in Computer Science, Specialization in Data Science and Engineering
Manipal Institute of Technology
Bachelor's degree, Electronics and Communications Engineering
Business Science University
DS4B 101-R: Business Analysis With R
Udacity
Data Engineering Nanodegree
ABOUT UTKARSH TREHAN
I\'m passionate about using technology to solve real-world business problems — with depth and curiosity. My journey began with a Bachelor\'s in Electronics and Communication Engineering and evolved through a Master’s in Data Science and Engineering, enriched by self-driven learning and hands-on projects. While I may not be a deep specialist in a single niche, my broad technological toolkit doesn’t dilute my expertise — it empowers me to bridge domains, connect the dots, and engineer solutions that are both creative and efficient.I approach problems with an end-to-end mindset — investing time upfront to understand context deeply before writing a line of code. I thrive in collaborative environments, but I’m equally comfortable taking initiative and owning solutions from idea to deployment.My time in research helped shape this perspective. During my Master’s, I worked on projects that stretched the boundaries of applied AI. SpectraINET, a wavelet-based CNN model for hyperspectral image classification, achieved state-of-the-art performance and was later published. I also explored adversarial robustness and model interpretability through a Grad-CAM generalization project, which I presented at CVPR workshops. These experiences deepened my belief in building systems that are not only intelligent but also rigorous and explainable.In essence, I’m a data science professional with a perpetual thirst for learning — one who enjoys bridging research and application to create meaningful impact.
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.