Shobhit C.

Key Skills

Software

No software listed

Top Subject Matter

Large Language Model training and prompt engineering

AI dataset annotation and model evaluation

Top Data Types

Text

Image

Video

Top Task Types

Prompt + Response Writing (SFT)

Data Collection

RLHF

Transcription

Freelancer Overview

I have hands-on experience in AI training, data annotation, and AI response evaluation through my work with Outlier AI and Aligner AI. At Outlier AI, I worked on multiple rubric-based projects where I evaluated AI-generated responses for instruction-following, factual accuracy, reasoning quality, clarity, and safety. I created complex prompts to identify model weaknesses, analyzed edge cases, detected hallucinations, and provided detailed corrective feedback to improve overall model performance. At Aligner AI, I worked as an AI Trainer and Data Annotator, where I performed data labeling, transcription, quality assurance, and real-time AI evaluation tasks. I have completed over 200 tasks across 20+ projects, consistently maintaining high accuracy and attention to detail. My experience as a Chegg Subject Matter Expert in Advanced Mathematics and Physics also strengthened my analytical thinking, editorial precision, and ability to produce clear, high-quality explanations, which helps me excel in AI training and evaluation roles.

Labeling Experience

Outlier AI — LLM Trainer (Remote)

Don't discloseTextPrompt + Response Writing (SFT)

Trained Large Language Models by generating complex prompt sets to elicit and identify model errors. Provided concise feedback and corrected solutions to improve the model’s reasoning and response generation. Delivered results by completing many training tasks across diverse projects. • Created prompts aimed at catching and correcting specific failure modes • Iteratively refined model outputs using short targeted corrections • Evaluated response quality during training task completion • Improved reasoning quality through structured feedback cycles

2025 - Present

Aligner AI — AI Trainer / Data Annotator (Remote)

Don't discloseTextData Collection

Annotated and transcribed data rows to improve overall AI accuracy and dataset quality. Conducted live interactions with AI models to test performance and supply real-time corrective guidance. Enriched datasets with relevant numerical data to improve contextual authenticity of AI responses. • Meticulously annotated data rows and transcribed required fields • Participated in live model evaluation sessions with corrective responses • Added numerical details to support more genuine context in outputs • Improved dataset usability for downstream AI training

2024 - Present

Education

Work History