Ranit Nandi

Key Skills

Software

Labelbox

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Document

Image

Text

Top Label Types

Text Summarization

RLHF

Fine Tuning

Evaluation Rating

Prompt Response Writing SFT

Freelancer Overview

I am an experienced GenAI Operations and Quality Manager with extensive involvement in training and evaluating large language models through RLHF, supervised fine-tuning (SFT), and agentic data workflows. At Turing, I lead AI data operations and quality calibration for globally distributed teams, ensuring high-quality datasets aligned with strict client standards and real-world deployment requirements. I have mentored and calibrated over 100 trainers and team leads, reviewed and optimized AI-generated outputs, and helped maintain consistency, accuracy, and scalability across large-scale LLM training pipelines. My expertise includes data annotation quality assurance, prompt and response evaluation, guideline development, workflow optimization, and performance monitoring using metrics and dashboards. I have collaborated on enterprise-level AI initiatives, contributed to improving training efficiency and dataset reliability, and developed automation and documentation to streamline complex data processes. With a strong technical background in software engineering, C++, and full-stack development, I bring both operational leadership and technical depth to AI training, evaluation, and quality management.

Labeling Experience

LLM Training Data Quality Calibration, RLHF & SFT Evaluation

LabelboxTextText SummarizationRLHF

Worked on large-scale training and evaluation of Large Language Models (LLMs) using Reinforcement Learning from Human Feedback (RLHF) and Supervised Fine-Tuning (SFT). Evaluated and labeled AI-generated responses for correctness, reasoning quality, instruction adherence, factual accuracy, and safety compliance. Performed prompt-response analysis, quality rating, and response correction to improve overall model performance and reliability. Led calibration and quality assurance efforts for a distributed team of 100+ trainers, ensuring consistency and adherence to strict client guidelines. Reviewed annotated datasets, identified quality gaps, and provided actionable feedback to improve annotation accuracy. Contributed to workflow optimization, guideline development, and performance tracking using dashboards and operational metrics. This work supported production-level GenAI systems used in real-world enterprise applications.

Worked on large-scale training and evaluation of Large Language Models (LLMs) using Reinforcement Learning from Human Feedback (RLHF) and Supervised Fine-Tuning (SFT). Evaluated and labeled AI-generated responses for correctness, reasoning quality, instruction adherence, factual accuracy, and safety compliance. Performed prompt-response analysis, quality rating, and response correction to improve overall model performance and reliability. Led calibration and quality assurance efforts for a distributed team of 100+ trainers, ensuring consistency and adherence to strict client guidelines. Reviewed annotated datasets, identified quality gaps, and provided actionable feedback to improve annotation accuracy. Contributed to workflow optimization, guideline development, and performance tracking using dashboards and operational metrics. This work supported production-level GenAI systems used in real-world enterprise applications.

2024 - 2025

Education

Work History