Hello, I'm Mohammed Asif Sahadh.

I'm currently pursuing my Master's in Data Science at Vellore Institute of Technology, Andhra Pradesh, while actively developing expertise in the fields of Large Language Models, Natural Language Processing, Deep Learning, and Machine Learning.

My interest for Data Science began during the second year of my bachelor's degree, when I took a course titled "Introduction to R Programming." The course was essentially an introduction to data analysis using R. However, what intrigued me was realizing technology's potential to convert seemingly mundane data files into something of immense benefit.

However, my interest for Artificial Intelligence and Machine Learning developed more recently. After my bachelor's, a senior of mine reached out to me, and, long story short, he generously offered to help me better understand and learn the field of AI and ML. Under his guidance, I was able to deepen my knowledge and advance in the field. The roadmap covered everything from learning the fundamentals of AI mathematics to implementing neural network architectures and understanding transformers. Now, I am on my own journey of learning and progressing in the field.

Experience

Data Science Intern | ExactSpace

November 2024 - Present | Bangalore, India

  • Building optimisation pipelines and simulations for industrial processes to analyse performance, recommend setpoints, and improve operational efficiency using historical and real-time sensor data.
  • Developed closed-loop control pipelines combining custom-made PIDs, multi-model ANN frameworks, transformer models, and solver-based optimisation to support WHRS (Waste Heat Recovery System) optimisation.
  • Built merit order optimisation models to determine optimal load distribution under varying environmental and plant conditions to support adaptive and highly efficient plant operation.
AI Intern | Amygda

June 2025 - September 2025 | England, UK

  • Researched & developed reliable Agents, Retrieval-Augmented Generation (RAG) systems & solutions based on Generative AI & NLP for deployment in the transportation sector.
  • Designed and implemented 2 pipelines incorporating Vision-Language Models (VLMs), including a RAG pipeline that answers user queries from domain-specific imagery across 6 volumes of engineering documentation.
  • Enhanced an existing Time Series Agent by making user query retrieval more deterministic while reducing token consumption by ~35%, lowering operational costs.
  • Explored & evaluated several open-source VLMs optimized for local deployment, carefully balancing performance & resource trade-offs.
  • Improved the existing agentic framework using LangGraph by integrating Human-in-the-Loop (HITL) only when essential, thereby increasing response accuracy.
Python Tutor | Cybeorg Education Technology Pvt. Ltd.

February 2025 - April 2025 | Delhi, India

  • Taught Python to young minds through interactive online sessions, while making sure that they acquire hands-on coding and problem-solving experience.
I.T Support | Techinsight Sdn. Bhd.

September 2022 - December 2022 | Kuala Lumpur, Malaysia

  • Contributed to the development of 2 distinct websites for the company, utilizing HTML and CSS.
  • Presented a mobile application developed by the company to a team with over 20 members.
  • Conducted thorough user acceptance testing of mobile applications developed by the company for a positive user experience.

Publications

Efficient Parameter Estimation by Fine Tuning DistilBERT for Synthetic Text Detection

IEEE | Feb 23, 2026

This study presents a lightweight text classification model designed to distinguish human-written content from text generated by large language models. Using transfer learning on a large and diverse dataset of 1.1 million samples from over 62 LLMs, the approach applies layer selective fine tuning to DistilBERT. This reduces computational cost while maintaining strong generalization and avoiding overfitting. The resulting model achieves high performance, including 97.10% accuracy, strong F1 score, excellent AUROC, and a low false negative rate.

Education

M.Sc Data Science | Vellore Institute of Technology, Andhra Pradesh

September 2024 - Present | Andhra Pradesh, India

B.Sc Computer Science (Hons) (Artificial Intelligence) | Asia Pacific University of Technology and Innovation

December 2020 - December 2023 | Kuala Lumpur, Malaysia

High School Diploma (Mathematics and Computer Science) | International Indian School, Jeddah

April 2018 - March 2020 | Jeddah, Saudi Arabia

Clubs & Societies

Deep Learning Team Co-Lead | The Machine Learning Club VIT-AP

November 2024 - Present | Andhra Pradesh, India

  • Lead a research project with a team of 5, focusing on non-convex optimization techniques to improve algorithmic efficiency and performance in machine learning applications.
  • Contributed to the development and testing of a computer vision-based gaming event in Minecraft.