Praful S.

Key Skills

Software

Other

Top Subject Matter

LLM quality assurance

conversational AI evaluation

safety & moderation

Top Data Types

Text

Document

Top Task Types

RLHF

Entity (NER) Classification

Classification

Freelancer Overview

AI Chat Evaluator & Model Trainer (Freelance/Remote Project Contract). Brings 7+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Internal, Proprietary Tooling, and Other. Education includes Bachelor of Arts, Lucknow University (2026) and Diploma in Pharmacy, Board of Technical Education Uttar Pradesh (BTEUP) (2025). AI-training focus includes data types such as Text and Document and labeling workflows including Evaluation, Rating, and Translation.

Labeling Experience

Data Quality Assurance Specialist - Remote Projects

TextRLHFEntity (NER) Classification

You applied data modeling and analytical methods to inspect and sanitize unstructured datasets prior to machine learning use. You performed verification testing on technical outputs and recorded procedural inconsistencies and edge cases. You authored clear technical and data documentation guidelines to support consistent remote annotation and cross-functional collaboration. • Modeled and analyzed raw unstructured datasets for readiness • Conducted verification tests and documented edge cases • Produced technical documentation and data guidelines for teams • Supported cross-functional remote review and quality alignment

2020 - Present

AI Chat Evaluator and Model Trainer - Freelance

TextRLHFEntity (NER) Classification

You evaluated and rated AI-generated text and conversational models for key quality dimensions including accuracy, relevance, helpfulness, and safety. You performed prompt-testing and output analysis to support large language model improvements within RLHF-style workflows. You conducted content moderation and data verification to filter out hallucinated or biased results and improve reliability. • Assessed factual accuracy, contextual relevance, and safety of outputs • Conducted prompt-engineering tests and analysis for model improvement • Performed moderation and verification to reduce hallucinations and bias • Collaborated in remote agile workflows to meet throughput and quality KPIs

2020 - Present

Technical Writer & Translator (English & Hindi) (Freelance)

OtherDocument

Translated technical, digital, and scientific articles between English and Hindi while maintaining strict context, tone, and grammatical accuracy. Used advanced internet research to cross-verify domain terminology and improve alignment of translated content with source meaning. Supported language localization for AI-facing or documentation-oriented materials through careful proofreading and terminology consistency. • Translated technical/scientific content English↔Hindi with style preservation • Cross-verified domain terminology using internet research • Ensured translation alignment and semantic consistency • Proofread/localized content for accurate communication

Not specified

Data Analytics & Quality Assurance Specialist (Remote Projects)

Text

Applied data modeling and analytical frameworks to inspect, sanitize, and prepare raw unstructured datasets for downstream machine learning integration. Executed verification testing on technical outputs, documenting procedural inconsistencies and identifying edge cases. Authored technical documentation and data documentation guidelines to align cross-functional remote annotation and QA efforts. • Inspected and sanitized unstructured datasets prior to ML use • Performed verification testing and reported inconsistencies and edge cases • Produced documentation to standardize remote QA/annotation • Supported cross-functional alignment for technical outputs

Not specified

AI Chat Evaluator & Model Trainer (Freelance/Remote Project Contract)

Text

Rated and evaluated AI-generated conversational outputs to improve model reliability across key quality dimensions including factual accuracy, relevance, helpfulness, and safety. Performed prompt-engineering testing and output analysis to support LLM training workflows using RLHF-style evaluation. Conducted content moderation and data verification by filtering hallucinated or biased results to optimize safety and performance. • Evaluated text and conversational model responses against defined criteria • Tested prompts and analyzed outputs for training readiness • Moderated content and verified data quality to reduce hallucinations • Worked in remote agile processes to meet throughput and QA KPIs

Not specified

Education

Work History