Chidera D.

Key Skills

Software

Labelbox

Toloka

Data Annotation Tech

Appen

Remotasks

Top Subject Matter

Medical imaging dataset QA for medical domain annotation

Multilingual African language NLP (English, Igbo, Nigerian Pidgin)

Speech and ASR data annotation for African languages/dialects

Top Data Types

Text

Audio

Image

3D Sensor

Document

Top Task Types

Transcription

RLHF

Classification

Object Detection

Entity (NER) Classification

Bounding Box

Freelancer Overview

Image & Text Annotation Lead (QA) | Labelbox — AI Startup Contract. Brings 4+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Labelbox, Toloka, and N. Education includes Bachelor of Medical Laboratory Science, University of Nigeria, Nsukka (2025) and Project Management Professional Certificate, Google / Coursera (2026). AI-training focus includes data types such as Medical, DICOM, and Text and labeling workflows including Evaluation, Rating, and Translation.

Labeling Experience

AI Training Data Contributor & RLHF Evaluator | DataAnnotation.tech

Data Annotation TechTextRLHF

Evaluated and rewrote 900+ prompt-response pairs for LLM RLHF training. Assessed outputs for factual accuracy, tone, helpfulness, and instruction adherence, then rewrote 400+ low-quality responses to improve clarity and contextual correctness. Maintained a 91% evaluator agreement score with expert reviewers and identified systematic model failure patterns to refine prompt guidelines. • RLHF output evaluation (accuracy/tone/helpfulness/instruction) • Prompt-response rewriting and style/tone improvements • Inter-evaluator agreement calibration • Failure-mode analysis and edge-case escalation

2023 - 2024

Research & Annotation Participant | Prolific

TextEntity (NER) Classification

Participated in 35+ AI research annotation studies covering commonsense reasoning, visual question answering, and cultural bias detection. Selected for invite-only African language NLP studies requiring native Igbo speakers and maintained a 100% approval rate with zero rejected submissions. Followed study-specific labeling guidelines to support research dataset creation and evaluation. • Commonsense reasoning annotation support • Visual question answering related annotation • Cultural bias detection labeling • Native Igbo-focused research annotation

2023 - 2023

Image & Text Annotation Lead (QA) | Labelbox — AI Startup Contract

Labelbox

Led quality review operations for a 5-member annotation team on a medical imaging dataset, ensuring guideline compliance. Reduced annotation error rates from 8% to below 3% by creating a shared edge-case reference guide adopted by the team. Reviewed 300+ annotated images weekly, resolving inter-annotator disagreements and documenting rationales. • Quality assurance and community moderation • Edge-case escalation and guideline enforcement • Consensus building for disagreement resolution • QC documentation and feedback loop maintenance

2023 - 2023

Audio & Text Annotation Contributor | Defined.ai

AudioTranscription

Transcribed and annotated 180+ hours of Nigerian English and Pidgin speech data for ASR model training. Kept word error rate below 3% through careful transcription quality controls. Conducted native-speaker validation for Igbo speech datasets to ensure phonological accuracy and labeled speaker intent and emotional tone across 1,100+ conversational audio clips. • ASR transcription and annotation • Native-speaker phonological validation (Igbo/Pidgin) • Speaker intent labeling • Emotion tone labeling for dialogue understanding

2022 - 2023

Multilingual Annotation Specialist | Yandex Toloka

TolokaText

Completed 2,800+ multilingual text classification and translation-quality tasks in English, Igbo, and Nigerian Pidgin for low-resource African language NLP. Maintained an average quality score of 4.8/5.0 and proactively identified systematic labeling inconsistencies before deployment. Applied cultural and linguistic expertise to flag mistranslations and misclassifications that monolingual annotators could not detect. • Multilingual translation quality assessment • Text classification labeling • Identification and reporting of labeling inconsistencies • Cultural/linguistic validation and error flagging

2022 - 2023

Show more

Education

Work History