Suraj Kalidindi Email & Phone Number

• Developed a more robust and accurate pipeline for product sentiment analysis using LLM prompting and aspect‑based sentiment analysis.
• Used Python with Langchain, Huggingface, and Claude’s API among other tools to create the pipeline, and did extensive research to find state of the art NLP methods for sentiment analysis.

Swarthmore, Pennsylvania, United States

• Lead the computational analysis of a significant Neuro‑toxicology study using Machine learning and Computer vision techniques. Our study employs various supervised, unsupervised, and semi‑supervised models to analyze.
• Developed Python scripts independently from scratch to extract useful features from video data ‑ these extracted features are used to trainclassification models that help us map specific chemicals to specific worm.
• Employed Computer vision and statistical techniques to crop videos, isolate foreground objects(the worm), and extract informative featuresfrom a large dataset of noisy videos with varying lighting conditions and quality.
• Utilize a diverse range of models such as LSTMs, CNNs (for classification tasks), Random Forests, XGBoost, T‑SNE, PCA, UMAP, and other unsupervised and semi‑supervised learning methods for in‑depth image analysis.
• Applied computer vision techniques extensively

Birmingham, Alabama

• Worked on a research project with my professor titled “Understanding the Impact of Social Determinants of Health on COVID-19 Outbreak.”
• Created functions that convert large heterogenous Covid 19 datasets to usable feature vectors for my professor’s research using various methods (Normalization, One-hot encoding, T-SNE, PCA, Isometric mapping).
• Contributed to our team achieving second place in the American Heart Association COVID-19 Data Challenge 2020.
• Completed Stanford’s Machine Learning course on Coursera, and the introductory machine learning challenge on kaggle.com.

Hyderabad, Telangana, India