Qasim I.

Key Skills

Software

Don't disclose

Top Subject Matter

LLM code & prompt evaluation (Python/C++)

Machine learning dataset preparation and model output evaluation (churn classification)

Top Data Types

Text

Top Task Types

Classification

Tracking

RLHF

Entity (NER) Classification

Freelancer Overview

AI Code & Prompt Evaluation — Independent Research (AI training/data feedback style work). Brings 2+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Don't disclose, Scikit-learn, and Pandas. Education includes Bachelor of Science, FAST-NUCES (2027) and Higher Secondary School Certificate, Government College Civil Lines (2023). AI-training focus includes data types such as Text, Computer Code, and Programming and labeling workflows including Evaluation, Rating, and Classification.

Labeling Experience

AI Code & Prompt Evaluation - Independent Research

ClassificationTracking

No description provided.

2025 - Present

AI Code & Prompt Evaluation — Independent Research (AI training/data feedback style work)

Don't discloseText

Reviewed and rated LLM-generated code responses for correctness, clarity, and instruction-following across Python and C++. Evaluated model outputs for logical errors, edge-case handling, and adherence to best practices, producing structured feedback labels. Annotated natural language prompts and model responses for relevance, helpfulness, and safety alignment. • Assessed code quality and explanatory clarity • Checked logical consistency and edge-case coverage • Labeled prompt/response pairs for relevance/helpfulness/safety • Provided structured feedback for model improvement

2025 - Present

Customer Churn Prediction Model — Python / Machine Learning

Built a customer churn prediction pipeline by cleaning and preprocessing a real-world Kaggle dataset for binary classification. Trained and evaluated a churn model using Scikit-learn with tuning for precision and recall. Although not a human annotation role, the work involved creating ground-truth quality through dataset cleaning and assessing model output performance for downstream labeling utility. • Performed dataset cleaning and preprocessing • Trained binary classification models and tuned metrics • Evaluated model predictions for quality • Deployed inference via a Streamlit app to validate outputs

2024 - 2024

Education

Work History