Sreeya M.

Key Skills

Software

Other

Mercor

Top Subject Matter

Large Language Model (Shopping Assistant) dataset curation and LLM training

RLHF evaluation of Hindi LLM responses and prompt-output correctness

Top Data Types

Text

Top Task Types

Prompt + Response Writing (SFT)

RLHF

Freelancer Overview

Data Associate (AI Trainer – LLM), Amazon | NILE (Shopping Assistant). Professional background includes roles such as Data Associate (AI Trainer – LLM) and Independent Contractor (RLHF Evaluation Specialist). Core strengths include Other and Mercor. Education includes Bachelor of Technology, Marri Laxman Reddy Institute of Engineering and Technology (2021). AI-training focus includes data types such as Text and labeling workflows including Prompt + Response Writing (SFT) and RLHF.

Labeling Experience

Independent Contractor — Hindi RLHF Generalist, Mercor

MercorTextRLHF

Evaluated and improved AI decision-making workflows by analyzing prompts and AI-generated outputs for structural correctness, clarity, and logical coherence. Reviewed complex logic chains to identify flaws in model reasoning and provide feedback to optimize RLHF pipelines. Enforced evaluation principles including correctness, conciseness, lawfulness, and harmlessness to reduce contradictions and bias. • Performed structured evaluations and feedback on logic chain reasoning • Assessed verbosity, professional neutrality, and readability of outputs • Identified and flagged logical contradictions, bias, and subjective moral judgments • Provided documented formatting and structure improvement recommendations

2023 - 2024

Data Associate (AI Trainer – LLM), Amazon | NILE (Shopping Assistant)

OtherTextPrompt + Response Writing (SFT)

Authored structured reasoning traces and produced high-quality training data for the NILE shopping-assistant LLM. Validated prompts and responses across 15+ complex workflows to ensure quality assurance and enterprise alignment. Evaluated LLM decision-making workflows for correctness and effective handling of product queries and customer reviews. • Created structured QA pairs and explanatory responses for training • Implemented logic-driven “Wh” question flow and irrelevant-query detection patterns • Documented step-by-step decision-making processes to extract domain-level content patterns • Maintained data quality under evolving project guidelines and tight timelines

2023 - 2024

Independent Contractor (RLHF Evaluation Specialist) - Mercor

TextRLHF

Independent Contractor providing reinforcement learning evaluation support by reviewing prompts and AI-generated outputs for structural and logical correctness. Conducted reviews of complex logic chains to identify reasoning flaws and deliver actionable feedback to strengthen RLHF pipelines. Applied core evaluation principles such as correctness, conciseness, lawfulness, and harmlessness while documenting formatting improvements for readability. • Evaluated AI decision-making workflows by assessing prompt-output pairs for clarity and logical coherence. • Reviewed logic chains to detect reasoning errors and support RLHF reinforcement feedback cycles. • Enforced evaluation criteria to reduce contradictions, bias, and subjective moral judgments. • Documented formatting and structural revisions to keep outputs professionally neutral and properly verbose.

Not specified

Education

Work History