Role Overview

• Work on advanced AI data training and evaluation projects
• Review AI-generated analytical responses and R code
• Create high-quality R-based solutions with detailed reasoning

Key Responsibilities

• Review AI-generated outputs for:
  • Correctness, clarity, and requirement adherence
• Identify issues in:
  • Statistical methods and data-wrangling workflows
• Validate analytical results and interpretations
• Develop expert-level explanations and model solutions
• Apply best practices in R programming and data analysis
• Compare and rank multiple AI-generated responses
• Fact-check technical and statistical accuracy
• Improve AI model performance, reliability, and clarity
• Create structured prompts and analytical responses
• Detect inaccuracies, inconsistencies, and biases

Required Qualifications

• Bachelor’s degree in:
  • Statistics / Mathematics / Computer Science / related field
• 2+ years of hands-on experience with R

Technical Skills

• Strong proficiency in:
  • Data wrangling and functional programming in R
  • Writing reusable code and building packages
• Solid understanding of:
  • Regression, statistical inference, and model validation
• Experience with:
  • End-to-end data analysis (cleaning, EDA, modeling, visualization)
• Familiarity with:
  • tidyverse, data.table, ggplot2
• Ability to select appropriate tools for analytical tasks

Professional Experience

• Background in roles such as:
  • Data Scientist
  • Statistician
  • Quantitative Analyst

Additional Requirements

• Experience using:
  • AI coding tools / large language models
• Excellent English writing and communication skills
• Strong analytical thinking and attention to detail

Preferred (Good to Have)

• Experience in:
  • AI data training or model evaluation
  • Reviewing AI-generated analytical content

What You’ll Do in This Role

• Create structured AI training content (prompts & responses)
• Evaluate and improve AI-generated analytical outputs
• Enhance model accuracy, clarity, and reliability
• Ensure correctness across diverse data and statistical use cases