AI Research Data Science Visalakshi Iyer
AI Research Data Science Visalakshi Iyer
Accomplished Data Scientist with proven ability to develop advanced systems, translate foundational project requirements into technical design,
work efficiently to cater to problem resolution as-in-when they show up and work with cross-functional teams. Seeking a challenging role in
initiatives that deliver cutting-edge advancements in data science
EXPERIENCE
Masters in Data Science, University of Arizona (GPA: 3.8) Aug '23 — May '25
Coursework: Machine Learning, Deep Learning, Natural Language Processing, Data Analysis, ML Gradient Optimization, Database
Management
Bachelors in Information & Technology, L. D. College of Engineering (GPA: 9.17) Jul '18 — Jun '22
Coursework: Data Structures and Algorithms, Web Development, Computer Networks, Operating Systems
PROJECTS
Scientific Fact Verification using CoT and In-Context Learning, LLM, GPT-4o Nov '20
Developed a verification system using Large Language Models to test their ability to perform simulation-based analysis on the scientific
claim dataset — SciFact.
Implemented the model's Python simulations on scientific phenomena, achieving a 58% success rate in claim verification.
Common Sense Reasoning, Fine Tuning, Zero-shot, Transformer BERT, GPT Nov '20
Analyzed zero-shot vs. fine-tuned model performance on common sense reasoning tasks (Riddle-Sense, Causal-Judgement, Odd-One-Out)
from the BIG-Bench dataset.
Fine-tuned DeBERTa-v3-base on the CommonsenseQA dataset, improving accuracy by 16% in reasoning tasks.
Compared GPT-4o, GPT-3.5, CommandR, and DeBERTa performance, highlighting the impact of fine-tuning on reasoning generalization.
Music Generation using Deep Learning, Python, RNN LSTM, music21, Streamlit Mar '22
Developed a Neural Network that breaks down musical notes from mono-harmonic musical instruments and generated music based on the
feed.
Implemented LSTM network for generating baseline product that can be customized for market use cases.
SKILLS