Afolabi K.

Key Skills

Software

Labelbox

CVAT

Micro1

Top Subject Matter

LLM Evaluation

Instruction Following

Text QA/QC

Top Data Types

Text

Video

Image

Top Task Types

RLHF

Action Recognition

Text Generation

Evaluation/Rating

Transcription

Prompt + Response Writing (SFT)

Object Detection

Freelancer Overview

AI Trainer & Data Labeling Specialist (LLM/Instruction Adherence). Core strengths include Labelbox, CVAT, and Airtable. Education includes Bachelor of Science, N/A. AI-training focus includes data types such as Text and Video and labeling workflows including RLHF, Action Recognition, and Evaluation.

Labeling Experience

AI Agent Transcript Annotator & Validator

Text

I performed specialized data extraction and validation for AI agent call transcripts, focusing on field accuracy and intent classification. My tasks included systematic annotation, correction set generation, and quality rating for outbound AI voice response data. I documented misclassification patterns and provided actionable adjustments for continual model improvement. • Reviewed transcript fields for extraction accuracy and classification quality. • Generated annotated correction datasets to enhance extraction parameterization. • Identified model failure modes and labeled erroneous outputs accordingly. • Supported iterative improvements for agent-based AI systems via schema definition.

2023 - Present

Multi-Modal Data Annotation Specialist (Video Action/Caption Review)

CVATVideoAction Recognition

In this multi-modal annotation role, I validated model-generated video captions for atomicity, confirmed correct action boundaries, and tracked object and human movements. I edited and corrected labels regarding left/right hand usage and object interaction across diverse video data. The process involved detecting ambiguous or low-quality annotations and providing feedback for continual improvement. • Verified start/end frame alignment for video action boundaries. • Assessed caption atomicity so each label represented an independent action accurately. • Tracked visual objects and classified action sequences in human movement data. • Flagged ambiguous or incorrect labels for rework and documented systematic model failure cases.

2023 - Present

AI Trainer & Data Labeling Specialist (LLM/Instruction Adherence)

LabelboxTextRLHF

As an AI Trainer & Data Labeling Specialist, I executed RLHF preference ranking and instruction adherence evaluation for a variety of LLMs. I assessed model-generated outputs, identifying and correcting hallucinations and misclassifications across thousands of text responses. My responsibilities included quality rating, defining schema guardrails, and creating correction sets for improved extraction parameters. • Conducted preference ranking of text outputs by helpfulness, factuality, tone, and safety. • Audited instruction adherence and flagged omissions or format issues in LLM responses. • Identified and labeled hallucinations and misclassified assertions for correction and retraining. • Authored data schemas and documentation to support structured data validation.

2023 - Present

Education

Work History