Manato O.

Key Skills

Software

Don't disclose

Top Subject Matter

Single-cell imaging

drug response prediction

translational medicine

Top Data Types

Text

Image

Audio

Top Task Types

Object Detection

Classification

Freelancer Overview

Undergraduate Researcher at Caicedo Lab — object detection and predictive modeling (Mar 2025–Present). Brings 3+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include PyTorch, NumPy, and scikit-learn. Education includes Bachelor of Science, University of Wisconsin Madison (2028). AI-training focus includes data types such as Medical, DICOM, and Text and labeling workflows including Object Detection and Classification.

Labeling Experience

Undergraduate Researcher at Caicedo Lab — object detection and predictive modeling (Mar 2025–Present)

Object Detection

Conducted ML-driven object detection and feature extraction on single-cell imaging data to quantify cell morphology, motility, and phenotypic outcomes under different drug treatments. Built and validated predictive models by optimizing workflows on GPU resources and benchmarking across multiple cell lines. The work emphasized reproducibility and robust model performance for downstream translational medicine usage. • Processed JSON-formatted imaging outputs to extract and quantify cell features. • Built/used predictive modeling pipelines to forecast treatment-induced changes in cell shape and motility. • Performed GPU-based optimization and cross-validation for model and workflow validation. • Benchmarked models across cell lines (e.g., A549, HeLa) to ensure robustness.

2025 - Present

Word Connotation Classifier — neural network from scratch (October 2024–January 2025)

TextClassification

Trained a feedforward neural network to classify word sentiment as positive or negative using a supervised text dataset. Implemented core neural network training components (activation functions, backpropagation, and gradient descent) and used standard preprocessing and feature extraction for model input. Visualized learning behavior to support model optimization and understanding of sentiment analysis. • Used scikit-learn for preprocessing and feature extraction from word data. • Implemented neural network training from scratch with NumPy and Pandas. • Applied backpropagation and gradient descent to optimize model weights. • Used Matplotlib/Seaborn to visualize training performance and monitor convergence.

2024 - 2025

Stanford University Research Internship — Audio Separation Analysis (June 2023–August 2023)

AudioClassification

Co-authored an audio-source separation and instrument identification study using supervised learning on labeled audio samples. Implemented FFT-based audio analysis features and trained a TensorFlow neural network with regularization to classify instruments. The approach achieved moderate cross-instrument accuracy across a multi-instrument dataset. • Used Librosa and SciPy with FFT processing to prepare audio features. • Trained a 4-layer TensorFlow neural network with ReLU activations and dropout regularization. • Ran extensive training (1,000 epochs) using ~10,000 labeled audio samples. • Reported ~50% accuracy across 20 orchestral instruments for identification.

2023 - 2023

Parking Space Occupancy AI research paper — YOLOv5 training (May 2023–September 2023)

ImageObject Detection

Developed and trained a YOLOv5-based object detection model to classify parking spaces as occupied versus vacant using a custom labeled image dataset. Built a real-time computer vision workflow for multi-object detection in single images and demonstrated low-data, iterative training feasibility. The output supports practical smart-city and IoT parking management scenarios. • Trained YOLOv5 on a custom dataset of 135 labeled images for occupancy classification. • Implemented a real-time object detection pipeline to detect and classify parking spaces. • Used labeled-image data to support rapid iteration and practical deployment goals. • Evaluated feasibility for smart city/IoT parking management use cases.

2023 - 2023

Education

Work History