Aditya Mishra

Key Skills

Software

Appen

Datature

Google Cloud Vertex AI

Labelbox

CVAT

Scale AI

Prodigy

Top Subject Matter

Multi-modal AI (Computer Vision, NLP, Autonomous Systems)

Autonomous Vehicle Perception Data

NLP/Conversational AI and E-commerce Search

Top Data Types

Text

Audio

Image

3D Sensor

Top Task Types

RLHF

Fine Tuning

Segmentation

Entity Ner Classification

Freelancer Overview

Senior Data Labeling Analyst. Core strengths include Labelbox, CVAT, and VGG Image Annotator (VIA). AI-training focus includes data types such as Image, Text, and 3D Sensor and labeling workflows including RLHF, Segmentation, and Entity (NER) Classification.

Labeling Experience

Senior Data Labeling Analyst

LabelboxImageRLHF

I led the full annotation process for a major multi-modal AI initiative, prioritizing extremely high data accuracy. I executed RLHF by rigorously evaluating AI-generated responses across key safety and quality metrics. I partnered with ML engineers to refine guidelines and enhance model robustness. • Led end-to-end annotation for 200,000+ monthly data points • Conducted RLHF ranking on AI outputs for honesty, harmlessness, and helpfulness • Developed annotation guidelines to reduce edge-case errors • Curated datasets to address model performance gaps

I led the full annotation process for a major multi-modal AI initiative, prioritizing extremely high data accuracy. I executed RLHF by rigorously evaluating AI-generated responses across key safety and quality metrics. I partnered with ML engineers to refine guidelines and enhance model robustness. • Led end-to-end annotation for 200,000+ monthly data points • Conducted RLHF ranking on AI outputs for honesty, harmlessness, and helpfulness • Developed annotation guidelines to reduce edge-case errors • Curated datasets to address model performance gaps

2024 - Present

AI Mentor & Business Specialist (AI Training/Data Labeling)

TextRLHFFine Tuning

As a Business Specialist and AI Mentor, I acted as the primary interface guiding internal AI models to align outputs with high-level executive requirements. I reviewed and refined AI-generated communications for professionalism, accuracy, and persuasive effectiveness, ensuring alignment with industry-specific standards. I architected data labeling and alignment workflows, trained models on business process conventions, and created regional etiquette mapping for international applicability. • Transitioned AI outputs from academic to executive-ready formats by providing reinforcement learning with human feedback (RLHF). • Conducted tone and sentiment optimization of over 500 weekly AI-generated client texts with detailed feedback for improved LLM performance. • Developed global etiquette guides to train AI models in region-specific communications and workplace standards. • Audited financial/logical outputs, identifying and correcting errors to ensure model reliability and reduce hallucinations.

As a Business Specialist and AI Mentor, I acted as the primary interface guiding internal AI models to align outputs with high-level executive requirements. I reviewed and refined AI-generated communications for professionalism, accuracy, and persuasive effectiveness, ensuring alignment with industry-specific standards. I architected data labeling and alignment workflows, trained models on business process conventions, and created regional etiquette mapping for international applicability. • Transitioned AI outputs from academic to executive-ready formats by providing reinforcement learning with human feedback (RLHF). • Conducted tone and sentiment optimization of over 500 weekly AI-generated client texts with detailed feedback for improved LLM performance. • Developed global etiquette guides to train AI models in region-specific communications and workplace standards. • Audited financial/logical outputs, identifying and correcting errors to ensure model reliability and reduce hallucinations.

2021 - Present

AI Specialist

LabelboxImageSegmentation

I performed semantic segmentation and LiDAR labeling to prepare training datasets for autonomous vehicles. I was recognized for exceptional annotation quality and also led a QA process for a team of junior annotators. I used automation tools to streamline data conversion and sorting. • Labeled large-scale datasets for autonomous vehicle training • Managed team QA and provided feedback • Automated data sorting and conversion with Python • Achieved Master Annotator status with 99% IQS

I performed semantic segmentation and LiDAR labeling to prepare training datasets for autonomous vehicles. I was recognized for exceptional annotation quality and also led a QA process for a team of junior annotators. I used automation tools to streamline data conversion and sorting. • Labeled large-scale datasets for autonomous vehicle training • Managed team QA and provided feedback • Automated data sorting and conversion with Python • Achieved Master Annotator status with 99% IQS

2022 - 2024

Data labeler

LabelboxTextEntity Ner Classification

I executed various NLP data annotation tasks, focusing on intent classification, entity extraction, and chatbot optimization. I improved AI output by identifying toxic content and hallucinations during early-stage testing. I also handled image dataset categorization for e-commerce applications. • Labeled data for NLP intent and entity recognition • Annotated image datasets to boost search engine accuracy • Detected and flagged hallucinations and toxic outputs • Enhanced chatbot accuracy through labeled datasets

I executed various NLP data annotation tasks, focusing on intent classification, entity extraction, and chatbot optimization. I improved AI output by identifying toxic content and hallucinations during early-stage testing. I also handled image dataset categorization for e-commerce applications. • Labeled data for NLP intent and entity recognition • Annotated image datasets to boost search engine accuracy • Detected and flagged hallucinations and toxic outputs • Enhanced chatbot accuracy through labeled datasets

2020 - 2022

Education

Work History