Junaid S.

Key Skills

Software

Other

Top Subject Matter

Generalist and legal-domain AI model evaluation and prompt/response quality assessment

Legal Services & Contract Review

Regulatory Compliance & Risk Analysis

Top Data Types

Text

Document

Top Task Types

Fine-tuning

RLHF

Evaluation/Rating

Prompt + Response Writing (SFT)

Segmentation

Freelancer Overview

AI Prompt Writer / AI Model Evaluator (Outlier AI, Freelance/Contract, Remote). Brings 5+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Other and V7 Darwin. Education includes Bachelor of Laws, University of the West of England (2023) and Certificate of Higher Education, University of London (2022). AI-training focus includes data types such as Text and Document and labeling workflows including Evaluation, Rating, and Fine Tuning.

Labeling Experience

AI Prompt Writer / AI Model Evaluator (Outlier AI, Freelance/Contract, Remote)

OtherText

Performed AI model evaluation and fine-tuning support across generalist and legal tasks by assessing reasoning quality, accuracy, and instruction-following. Designed complex prompts to probe model weaknesses and compared outputs against structured task requirements. Reviewed and rewrote model responses to improve clarity, factual accuracy, structure, tone, and usefulness. • Developed evaluation rubrics covering reasoning quality, completeness, legal analysis, factual accuracy, and safety • Conducted response assessment to identify gaps in legal-focused issue analysis and drafting-style outputs • Collaborated with a remote global contributor network while maintaining confidentiality and meeting quality targets

2025 - 2025

AI Prompt Writer / AI Model Evaluator (Excluded from data labeling) - Outlier AI

Document

Contributed to AI model evaluation and fine-tuning initiatives focused on reasoning quality, accuracy, and instruction-following across generalist and legal tasks. Designed prompts for performance testing and reviewed generated responses for clarity, structure, factual accuracy, and usefulness. Developed and refined evaluation rubrics to measure legal analysis completeness and safety requirements, working within a confidential remote contributor network. • Created complex prompts to identify reasoning weaknesses • Rewrote AI outputs to improve factual accuracy and tone • Built evaluation rubrics for reasoning, completeness, and safety • Assessed model responses for legal and task-specific instruction adherence

2024

Education

Work History