cv

Bachelor

Indian Institute of Information Technology Kottayam

Computer Science (B.Tech with Honours)

Research Intern

Language Technology Group, University of Hamburg

Conducted research on embedding models and large language model (LLM) representations.

• Developed an optimal few-shot fine-tuning regime for topic modeling.
• Devised a novel pruning procedure for LLM-based embedding models, reducing model size by 21% with negligible performance drop.
• Published in RepL4NLP 2025. Internship was supported by the DAAD WISE scholarship.

Machine Learning Intern

Institute of Human Resource Development

Developed and deployed machine learning models for speech recognition, speaker identification, and face recognition.

• Worked with the Kerala Police Intelligence Department, leading a team of 15 to develop comprehensive AI systems for policing.
• Trained and deployed a state-of-the-art automatic speech recognition (ASR) model for Malayalam.
• Designed and implemented a Malayalam news extraction system using web scraping and OCR.
• Developed a scalable face recognition system optimized for fast inference.

Data Scientist Intern

Institute of Human Resource Development

Optimized web infrastructure and data-driven decision making.

• Rebuilt web stack to streamline employee workflows and replace outdated systems.
• Created visualizations and reports to communicate insights to internal stakeholders and government agencies.
• Collaborated with the procurement team to optimize purchasing decisions.

Sparsification for Model Merging

• Investigated the effects of sparsification on four representative model merging techniques with Dr. Suchithra M S as my Honours thesis.
• Demonstrated that sparsifying delta parameters significantly improves the performance of the merged multitask model.
• Established a systematic framework for integrating pruning into model merging workflows.
• Submitted to IEEE/ACM Transactions on Audio, Speech, and Language Processing.

Advocating for Character Error Rate in Automatic Speech Recognition

• Documented shortcomings of the commonly used word error rate metric for multilingual evaluation with Dr. Jesin James, Senior Lecturer at University of Auckland.
• Conducted multilingual surveys collecting human preferences among different ASR models.
• Calculated metric correlations, providing experimental evidence in favor of Character Error Rate.
• Published in NAACL Findings 2025.

Fisher Mask Nodes for Model Merging

• Developed a novel and compute-efficient model merging algorithm with Dr. Suchithra M S.
• Evaluated performance on various BERT family models, achieving a performance improvement of +6.5%.
• Achieved speedups between 57.4x and 321.7x.
• Published in LREC-COLING 2024.

Whisper Malayalam

• Trained ASR models on Malayalam speech data as part of the Huggingface Whisper Fine-Tune Community Sprint
• Elevated the medium-sized model's performance to the top of the leaderboard, making it the state-of-the-art solution for Malayalam ASR as evidenced by 30k total downloads and counting.

ICFOSS Malayalam Speech Corpus

• Collaborated with the International Centre for Free and Open Source Software (ICFOSS) on IMaSC - The ICFOSS Malayalam Speech Corpus, a 50-hour text-to-speech dataset.
• Supervised data collection, speaker recording, and quality control.
• Trained and evaluated multiple models, achieving an average MOS score of 4.50.

Data Augmentation for Automatic Voice Disorder Detection

• Evaluated data augmentation techniques for automatic voice disorder detection, focusing on leukoplakia with Dr Vrinda V Nair, Dean (Research) at APJ Abdul Kalam Technological University.
• Developed a custom data augmentation strategy that increased dataset size by 8x and achieved a 46.9% increase in accuracy.
• Published in a peer-reviewed journal.

DAAD WISE Scholarship

German Academic Exchange Service

Large Language Models Are Overparameterized Text Encoders

Proceedings of the 10th Workshop on Representation Learning for NLP (RepL4NLP-2025)

Advocating Character Error Rate for Multilingual ASR Evaluation

Findings of the Association for Computational Linguistics: NAACL 2025

Fisher Mask Nodes for Language Model Merging

Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024)

IMaSC -- ICFOSS Malayalam Speech Corpus

arXiv Preprint

Performance Enhancement of Deep Neural Network Based Automatic Voice Disorder Detection System with Data Augmentation — A Case Study

Biomedical Engineering: Applications, Basis and Communications, Vol. 35

Memory Based Speech Duration Model using Exemplar Theoretic Approach

International Conference on Artificial Intelligence & Speech Technologies (AIST 2019)