Nitish

Key Skills

Software

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Text

Top Label Types

Prompt Response Writing SFT

Freelancer Overview

I am an AI Evaluation Specialist with hands-on experience in data annotation, prompt creation, and quality assurance for large language models. My background includes evaluating and refining AI-generated outputs for logic, clarity, and factual accuracy, as well as designing and reviewing complex reasoning tasks. I am skilled in structured data interpretation, scenario-based evaluation, and ensuring linguistic and cultural accuracy in both English and Hindi. My expertise in prompt engineering, content quality checks, and multi-step task analysis enables me to contribute to the development of reliable and ethical AI systems. I excel in asynchronous, remote environments and am passionate about improving the quality and safety of AI through detailed data labeling and annotation work.

Labeling Experience

AI Tutor

Scale AITextPrompt Response Writing SFT

The Pangolin SFT project on Outlier is a high-level multimodal AI training initiative focused on enhancing the visual reasoning and instruction-following capabilities of frontier Large Language Models. In this role, specialists source complex visual data and craft sophisticated, multi-turn prompts that challenge the AI to perform tasks like spatial reasoning, detailed image interpretation, and accurate Optical Character Recognition (OCR). By rigorously evaluating model outputs, correcting hallucinations, and ensuring strict adherence to logical constraints, contributors develop the "Golden Datasets" essential for the Supervised Fine-Tuning phase, ultimately teaching the AI to accurately "see" and analyze the world in a human-like context.

The Pangolin SFT project on Outlier is a high-level multimodal AI training initiative focused on enhancing the visual reasoning and instruction-following capabilities of frontier Large Language Models. In this role, specialists source complex visual data and craft sophisticated, multi-turn prompts that challenge the AI to perform tasks like spatial reasoning, detailed image interpretation, and accurate Optical Character Recognition (OCR). By rigorously evaluating model outputs, correcting hallucinations, and ensuring strict adherence to logical constraints, contributors develop the "Golden Datasets" essential for the Supervised Fine-Tuning phase, ultimately teaching the AI to accurately "see" and analyze the world in a human-like context.

2024 - 2025

Education

Work History