I am a Master's-qualified Data Science and AI professional based in Sweden with strong competence in end-to-end development, from prototyping using PyTorch and TensorFlow to deployment via Hugging Face and Streamlit. I have successfully developed production-grade NLP solutions with over 90% accuracy on BERT models and am passionate about implementing privacy-preserving systems such as Differential Privacy. Combining strong technical skills with a passion for analytical problem-solving, I tackle diverse data challenges efficiently. Throughout my career, I have gained experience in Named Entity Recognition, AI engineering in game development using Unreal Engine 5, and advanced privacy research involving federated learning and secure multiparty computation. I thrive in collaborative, innovative environments and continuously seek to apply my expertise in machine learning, NLP, and privacy-focused AI development to meaningful projects and products.

Kaliyaperumal

I am a Master's-qualified Data Science and AI professional based in Sweden with strong competence in end-to-end development, from prototyping using PyTorch and TensorFlow to deployment via Hugging Face and Streamlit. I have successfully developed production-grade NLP solutions with over 90% accuracy on BERT models and am passionate about implementing privacy-preserving systems such as Differential Privacy. Combining strong technical skills with a passion for analytical problem-solving, I tackle diverse data challenges efficiently. Throughout my career, I have gained experience in Named Entity Recognition, AI engineering in game development using Unreal Engine 5, and advanced privacy research involving federated learning and secure multiparty computation. I thrive in collaborative, innovative environments and continuously seek to apply my expertise in machine learning, NLP, and privacy-focused AI development to meaningful projects and products.

Available to hire

I am a Master’s-qualified Data Science and AI professional based in Sweden with strong competence in end-to-end development, from prototyping using PyTorch and TensorFlow to deployment via Hugging Face and Streamlit. I have successfully developed production-grade NLP solutions with over 90% accuracy on BERT models and am passionate about implementing privacy-preserving systems such as Differential Privacy. Combining strong technical skills with a passion for analytical problem-solving, I tackle diverse data challenges efficiently.

Throughout my career, I have gained experience in Named Entity Recognition, AI engineering in game development using Unreal Engine 5, and advanced privacy research involving federated learning and secure multiparty computation. I thrive in collaborative, innovative environments and continuously seek to apply my expertise in machine learning, NLP, and privacy-focused AI development to meaningful projects and products.

See more

Experience Level

Expert
Expert
Expert
Intermediate
Intermediate
Intermediate
Intermediate
Intermediate
Intermediate
Intermediate
Intermediate
See more

Language

English
Fluent
Swedish
Beginner

Work Experience

AI Engineer Internship at Niyata Infotech Pvt.
January 1, 2024 - Present
Developed a Named Entity Recognition (NER) system using BERT (Hugging Face) to extract food items, quantities, and customizations from unstructured orders. Fine-tuned BERT on custom datasets, optimized tokenization including input IDs and attention masks, and benchmarked performance against DeepSeek. Containerized the model and API using Docker for consistent deployment and automated CI/CD pipelines using Azure DevOps. Deployed scalable inference on Kubernetes, reducing manual processing by 70% through structured JSON outputs.
AI Developer Internship at Realdini Studios
December 31, 2023 - August 26, 2025
Designed NPC and enemy behavior using AI in Unreal Engine 5 including intelligent enemy patrol, chase, and attack mechanics. Used language models to simulate responsive and natural NPC dialogue and enhanced game immersion by refining environment interaction logic.
Master Thesis Researcher
December 31, 2024 - August 26, 2025
Conducted research exploring privacy-accuracy trade-offs in health data using Differential Privacy (DP) and Secure Multiparty Computation. Implemented Laplace, Gaussian, and log-normal DP mechanisms and analyzed federated versus centralized methods, concluding federated learning enhances privacy with minimal accuracy loss. Applied DP techniques to hypothesis testing balancing privacy and statistical power, demonstrating global DP as optimal for decentralized health data analysis.

Education

Master's degree at Luleā University of Technology
August 28, 2023 - September 30, 2024
Bachelor of Engineering-Computer Science Engineering - AI and ML at Annamalai University
August 5, 2019 - May 30, 2023
Master of Science Data Science at Luleå University of Technology
January 1, 2023 - January 1, 2024
Bachelor of Engineering Machine Learning at Annamalai University
January 11, 2030 - August 26, 2025

Qualifications

Data Science Job Simulation – Forage in collaboration with BCG
December 4, 2024 - August 26, 2025
Data Analytics and Visualization Job Simulation – Forage in Collaboration with Accenture North America
April 22, 2025 - August 26, 2025
Data Analysis with Python – IBM DA0101EN Certificate | Cognitive Class
January 11, 2030 - August 26, 2025

Industry Experience

Software & Internet, Healthcare, Gaming, Education