Bashar I.

Key Skills

Software

Don't disclose

Other

Top Subject Matter

Multilingual AI evaluation and ML dataset annotation

Sentiment labeling and risk-signal annotation for teacher feedback

Model output review and quality assurance for prediction systems

Top Data Types

Text

Document

Top Task Types

Emotion Recognition

Classification

Entity (NER) Classification

Tracking

RLHF

Freelancer Overview

AI Text Annotator & LLM Output Evaluator (Academic & Project-Based, Self-Directed). Brings 1+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Internal and Proprietary Tooling. Education includes Bachelor of Computer Science (Data Science), Albukhary International University (2023). AI-training focus includes data types such as Text and labeling workflows including Evaluation, Rating, and RLHF.

Labeling Experience

AI Text Annotator & LLM Output Evaluator (Academic & Project-Based, Self-Directed)

Text

Performed rubric-based evaluation and scoring of AI/ML model outputs to identify labeling errors, hallucinations, reasoning inconsistencies, and classification mismatches. Applied annotation guidelines to validate 100,000+ rows while maintaining consistency and full audit traceability across multi-reviewer workflows. Leveraged multilingual capability (native Arabic and fluent English) to assess contextual correctness, semantic errors, and cross-lingual accuracy in multilingual NLP outputs. • Risk-signal label and sentiment classification schema design for human-in-the-loop checks before stakeholder delivery. • Compared predictions against ground-truth labels to find systematic errors and edge cases. • Documented discrepancies in structured reports to guide model retraining cycles. • Conducted multi-reviewer quality assurance with structured feedback labeling and traceability.

2024 - Present

ML Data Engineer (Anomaly Detection Pipeline) - Independent

TextClassificationTracking

Architected an end-to-end ML data pipeline for system performance monitoring and automated anomaly detection. Replaced reactive manual inspection with always-on anomaly detection validated to achieve strong precision. Built statistical validation workflows, error analysis reporting, and monitoring dashboards to support maintenance handoff. • Designed continuous tracking mechanisms to support operational reliability • Implemented model monitoring for ongoing performance review • Created structured documentation for classification of errors and anomalies • Built validation routines to ensure dependable anomaly outputs over time

2026 - 2026

Predictive Maintenance & Automated Anomaly Detection System

OtherText

Architected an end-to-end ML data pipeline for always-on system performance monitoring using automated anomaly detection. Conducted statistical validation and error analysis reporting to support trustworthy evaluation of anomaly signals. Produced model monitoring dashboards to enable continuous performance tracking and handoff. • Always-on anomaly detection for system performance monitoring • Statistical validation workflows and error analysis reporting • Model monitoring dashboards for continuous evaluation • Automation replacing reactive manual inspection

2026 - 2026

RainToday ML Pipeline — Production-Grade AI Data Engineering

OtherText

Built a leakage-safe Scikit-Learn pipeline with automated seasonal and categorical feature engineering to ensure clean training versus inference behavior. Performed comprehensive model evaluation and error documentation to produce reliable performance assessments suitable for deployment. Implemented cross-validation and monitoring so that labeled/derived features and evaluation outcomes can be continuously tracked. • Automated seasonal and categorical transformations via ColumnTransformer • Compared Logistic Regression vs Random Forest using recall/precision metrics • Cross-validation, performance monitoring, and structured error documentation • Feature-importance analysis identifying top predictors for model readiness

2026 - 2026

ML Output Review & Quality Assurance | RainToday Prediction System

Don't discloseText

Evaluated model prediction outputs against ground-truth labels to identify systematic errors and edge cases in the RainToday prediction system. Documented error patterns and annotation discrepancies in structured review reports to guide model retraining cycles. Benchmarked validation performance using precision targets to verify output quality. • Performed QA of predictions by comparing against labeled ground truth. • Flagged recurring failure cases for retraining prioritization. • Produced structured documentation for model improvement iterations. • Supported quality assurance with quantitative precision measurement (83% benchmark).

2026 - 2026

Show more

Education

Work History