Imran U.

Key Skills

Software

Labelbox

Mindrift

Google Cloud Vertex AI

Data Annotation Tech

CVAT

Dataloop

iMerit

HiveMind

LightTag

Micro1

Toloka

V7 Labs

Supervisely

Top Subject Matter

LLM training data generation

preference tuning

and code/script evaluation for AI training pipelines.

Top Data Types

Text

Audio

Document

Top Task Types

RLHF

Question Answering

Data Collection

Classification

Segmentation

Prompt + Response Writing (SFT)

Evaluation/Rating

Fine-tuning

Text Summarization

Text Generation

Polygon

Bounding Box

Freelancer Overview

Senior AI Training Expert & Code Evaluator (AI training and evaluator; RLHF preference tuning and dataset-quality valida. Brings 3+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Labelbox and Other. Education includes Master of Science, University of Lagos (Unilag) and Bachelor of Science, Ahmadu Bello University (Abu). AI-training focus includes data types such as Computer Code, Programming, and Text and labeling workflows including RLHF, Evaluation, and Rating.

Labeling Experience

Senior AI Training Expert & Code Evaluator (AI training and evaluator; RLHF preference tuning and dataset-quality validation).

LabelboxRLHF

Trained and evaluated AI preference models by applying RLHF methodologies to score responses for truthfulness and alignment. Authored structured multi-step prompt-and-answer materials to support LLM fine-tuning workflows. Performed systematic review of AI-generated code scripts to prevent execution edge cases.• Created and authored logically complex prompt/response datasets for model fine-tuning.• Evaluated model preference profiles using truthfulness and alignment scoring (via JSON/XML/Markdown).• Reviewed Python and SQL generated scripts for syntax and logical correctness.• Partnered with engineering leads to refine annotation guidelines and reduce dataset anomalies by 20%.

2022 - Present

Systems Analyst & Database Verification Lead (NLP token annotation and dataset verification).

OtherText

Led database verification and data digitization work that included annotating large token blocks for NLP parsing models. Conducted regular SQL dataset checks to clean and validate unstructured data matrices for downstream automation. Ranked among top distributed workers based on output speed and validation performance metrics.• Migrated, audited, and validated large student credential and CGPA tables for reliable structured inputs.• Annotated token blocks to improve structural syntax and semantic accuracy for NLP parsing models.• Designed and executed recurring SQL checks to clean unstructured data matrices.• Achieved top-5% performance on speed and validation scoring in distributed worker settings.

2019 - 2021

Education

Work History