Akash Narayan

Key Skills

Software

Scale AI

Labelbox

OneForma

Internal/Proprietary Tooling

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Text

Top Label Types

RLHF

Fine Tuning

Evaluation Rating

Emotion Recognition

Function Calling

Transcription

Question Answering

Text Generation

Translation Localization

Freelancer Overview

I am an AI Model Evaluation Specialist with hands-on experience in data labeling, annotation, and quality control across SFT and RLHF projects. My background includes evaluating and correcting AI-generated text and code, conducting linguistic annotation, and transcribing and aligning audio data for multilingual speech datasets. I am skilled in using tools like Labelbox, Scale AI, OneForma, and Aligner, and I have experience with rubric-based evaluation, hallucination detection, and feedback writing to improve model outputs. My technical foundation in Python, data analysis, and AWS cloud services is complemented by strong linguistic abilities in English, Hindi, Hinglish, and Bhojpuri, allowing me to adapt content for diverse audiences and cultural contexts. I am comfortable working with evolving guidelines, maintaining high accuracy, and meeting productivity targets in fast-paced, human-in-the-loop AI workflows.

Labeling Experience

Speech Data Contributor

Internal Proprietary ToolingAudioTranscription

Transcribed spoken audio into text with accurate punctuation, speaker fidelity, and formatting. Aligned spoken and written transcripts for model training. Ensured cultural and linguistic correctness for Hindi-English content.

Transcribed spoken audio into text with accurate punctuation, speaker fidelity, and formatting. Aligned spoken and written transcripts for model training. Ensured cultural and linguistic correctness for Hindi-English content.

2025

Linguistic Contributor

OneformaTextQuestion AnsweringText Generation

Evaluated and corrected AI-generated linguistic outputs in English, Hindi, and Hinglish. Reviewed grammar, spelling, context, clarity, and cultural appropriateness. Applied linguistic style guidelines across multiple content domains and delivered outputs under tight deadlines.

Evaluated and corrected AI-generated linguistic outputs in English, Hindi, and Hinglish. Reviewed grammar, spelling, context, clarity, and cultural appropriateness. Applied linguistic style guidelines across multiple content domains and delivered outputs under tight deadlines.

2025

Linguistic Annotator & Audio Labeling

LabelboxAudioEmotion RecognitionEvaluation Rating

Aligned audio speech segments to transcripts for linguistic accuracy. Performed detailed consistency checks on speaker timing, word boundaries, phonetic alignment, and pacing. Contributed to annotation quality checks and accuracy improvements for downstream ASR training.

Aligned audio speech segments to transcripts for linguistic accuracy. Performed detailed consistency checks on speaker timing, word boundaries, phonetic alignment, and pacing. Contributed to annotation quality checks and accuracy improvements for downstream ASR training.

2025

AI Model Evaluation Specialist

Scale AITextRLHFFine Tuning

Evaluated and rated 2000+ AI-generated responses for correctness, relevance, hallucination, tone, and intent alignment. Provided structured feedback to improve model behavior and safety. Identified linguistic errors, hallucinations, and logical inconsistencies. Collaborated with distributed raters to maintain standardized quality metrics.

Evaluated and rated 2000+ AI-generated responses for correctness, relevance, hallucination, tone, and intent alignment. Provided structured feedback to improve model behavior and safety. Identified linguistic errors, hallucinations, and logical inconsistencies. Collaborated with distributed raters to maintain standardized quality metrics.

2024

Education

Work History