Devraaj K.

Key Skills

Software

Other

Top Subject Matter

AI model evaluation and prompt/pattern extraction for large language model tasks

Agricultural domain AI model evaluation for branding content

Legal Services & Contract Review

Top Data Types

Text

Document

Top Task Types

Evaluation/Rating

Transcription

Text Summarization

Object Detection

Text Generation

Question Answering

Freelancer Overview

Generalist and Marketing Strategist — Outlier AI (AI training contributor). Brings 8+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include MultiMango and Other. Education includes Associate in Media and Communications, Lasalle College (2025) and Bachelor of Science, R.K. University (2020). AI-training focus includes data types such as Text and labeling workflows including Evaluation and Rating.

Labeling Experience

Generalist and Marketing Strategist — Outlier AI (AI training contributor)

Text

Executed pattern extraction across large unstructured datasets to identify recurring structures, edge cases, and failure modes. Produced original reference content used to train and calibrate model outputs, achieving over 80% accuracy across tasks. Designed prompts to test model behavior and refined weaknesses related to reasoning, factual accuracy, and instruction following. • Conducted iterative task cycles for pattern extraction and documentation • Informed correction workflows and provided feedback that influenced training UI and evaluation • Synthesized findings to enable faster onboarding and consistent evaluation across contributors • Improved downstream model evaluation quality through targeted analysis

2026 - 2026

Marketing & Brand Manager — FarmlY’s Agri. Tech (AI model evaluation)

OtherText

Evaluated a Claude Sonnet-based AI model for branding and agricultural-domain responses related to crops, soil, and farming knowledge. Assessed response accuracy to ensure outputs aligned with an agriculture-focused tool’s expected knowledge and behavior. Provided feedback that supported model calibration for domain-appropriate content. • Reviewed AI output quality against domain expectations • Validated factual accuracy for agricultural terminology and guidance • Supported refinement of model behavior for marketing/branding use cases • Contributed to assessment workflows for downstream AI usage

2024 - 2025

Education

Work History