Musi Calori

Key Skills

Software

Appen

Mercor

OneForma

Other

Remotasks

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Computer Code Programming

Text

Video

Top Label Types

Classification

Data Collection

Prompt Response Writing SFT

Text Generation

Freelancer Overview

I am a data analyst and scientist with hands-on experience in building end-to-end machine learning projects, focusing on transforming raw data into actionable insights and robust AI solutions. My expertise includes data cleaning, annotation, and wrangling, as well as exploratory data analysis, model evaluation, and handling imbalanced datasets—skills crucial for high-quality data labeling and AI training data preparation. I am proficient in Python (Pandas, NumPy, Scikit-learn), SQL, Excel, and visualization tools like Tableau and Jupyter Notebook. Through projects such as an AI-driven financial health monitoring system, predictive analytics for community beekeeping, and customer churn prediction, I have worked extensively with real-world datasets, applying both supervised and unsupervised learning techniques. I am passionate about ensuring data quality and integrity, and I excel at translating complex data into clear, strategic insights for AI model development. I am eager to contribute to projects that require meticulous data annotation and training data management, continuously learning and adapting to new challenges in the AI and data science space.

Labeling Experience

collection of video

AppenVideoData Collection

Collecting real crying audio data from infants, including cries under different conditions such as hunger, during sleep, and after waking up. Maximum 1 hour of valid audio per infant. Each individual crying segment must be longer than 300 ms. - Continuous recording is allowed. Background noises (including non-crying infant sounds, adult crying, adult speech, adult laughter, crying/laughing/speech of older children, cat meows, dog barks, household appliances, vehicle sounds, traffic noise, etc.) are acceptable and will not affect data submission. - Audio must not be clipped (avoid clipping caused by being too close to the microphone or excessive volume). - Recording distance: 0.5–2 meters from the infant.

Collecting real crying audio data from infants, including cries under different conditions such as hunger, during sleep, and after waking up. Maximum 1 hour of valid audio per infant. Each individual crying segment must be longer than 300 ms. - Continuous recording is allowed. Background noises (including non-crying infant sounds, adult crying, adult speech, adult laughter, crying/laughing/speech of older children, cat meows, dog barks, household appliances, vehicle sounds, traffic noise, etc.) are acceptable and will not affect data submission. - Audio must not be clipped (avoid clipping caused by being too close to the microphone or excessive volume). - Recording distance: 0.5–2 meters from the infant.

2025 - 2025

LLMs training

OtherTextPrompt Response Writing SFT

giving response analysis to LLMs generated responses

giving response analysis to LLMs generated responses

2025 - 2025

Education

Work History