Randhir G.

Key Skills

Software

Other

Top Subject Matter

AI safety

RLHF evaluation

conversational quality

Top Data Types

Text

Document

Top Task Types

RLHF

Evaluation/Rating

Prompt + Response Writing (SFT)

Freelancer Overview

AI Evaluator / RLHF & Safety Reviewer (AI-generated response quality and preference evaluation). Brings 11+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Other. Education includes Master of Business Administration, Mumbai University (2002) and Bachelor of Engineering, Mumbai University (1998). AI-training focus includes data types such as Text and labeling workflows including RLHF.

Labeling Experience

AI Evaluator / RLHF & Safety Reviewer (AI-generated response quality and preference evaluation)

OtherTextRLHF

Performed comparative AI response evaluation to support RLHF-style human preference learning and iterative model improvement. Assessed prompt-response quality using rubric-based judgments for safety, policy alignment, hallucination likelihood, and conversational calibration. Produced structured, evidence-based reviewer notes and rankings to reflect human-like preferences and detect subtle reasoning issues. • Response A vs B comparative ranking and preference scoring. • Safety, privacy, and governance reasoning checks including hallucination/misinformation detection. • Emotional calibration and tone evaluation for conversational quality. • Ambiguity and tradeoff analysis with escalation and justification writing.

2016 - Present

Education

Work History