For employers

Hire this AI Trainer

Sign in or create an account to invite AI Trainers to your job.

Invite to Job
D
Daniel Ulokaji

Daniel Ulokaji

AI Data Annotator - IT

Nigeria flagOGUN, Nigeria
$5.00/hrEntry LevelLabel StudioSurge AIArgilla

Key Skills

Software

Label StudioLabel Studio
Surge AISurge AI
ArgillaArgilla
EncordEncord

Top Subject Matter

Technology
Finance
E-Commerce

Top Data Types

TextText

Top Task Types

RLHFRLHF
TranscriptionTranscription

Freelancer Overview

Education includes Bachelor of Engineering, University of Benin (2025). Experience annotation process for AI Training Data. From raw data collection all the way through successfully fine-tuning high performance models. Advanced from single object bounding box labeling into instruction-tuning dataset generation and RLHFAlignment. Auditing multi-turn conversations where the objective is not only accuracy but conversation-level safety and perceived "helpfulness" through thousands of tokens. Technical experience performing and managing detailed labeling operations as well as prompt engineering and synthetic data generation to augment sparse training data sets. What makes me different is that I marry data hygiene with thoughtful annotation strategy. I'm not all about high volume labeling, I speak your language of converting unstructured messy data into defined schema structures (JSON/XML) for finetuning. Designed and executed pipelines for cleaning noisy data that involved both automated regex filtering and human qualitative analysis. Cutting significant noise has greatly increased recall. Proficient at "red teaming" evaluation of where a model will hallucinate / fail safety checks, enabling me to label only the counterfactuals that actually harden models vs. the most obvious positive cases.

Entry LevelEnglish

Labeling Experience

Safety & Preference Alignment Labeling for Conversational LLM Fine-Tuning

TextRLHF
I worked on a large scale Reinforcement Learning from Human Feedback (RLHF) pipeline to fine-tune a generative AI assistant to be safer and follow instructions better as a Tier 2 annotator and auditor. This involved ranking preferences instead of binary ranking as well as "red-teaming". Major Accomplishments & Skills Showcased Multi-Turn Context Reviews: Completed reviews of conversations with more than 10 dialogue turns to check that the model remained consistent and didn't provide contradictory/unsafe responses over longer conversations. Adversarial Red-Teaming & Edge Case Labeling: Human-written edge-case prompts created to elicit hallucinations/bias/too-much refusal. Counterfactual responses were written to train the model why a particular response was harmful instead of just labeling it "bad". Granular Preference Ranking: Performed side-by-side preference labeling with granular written explanations. Prioritized tone, succinctness, and strict adherence to safety guidelines over objective correctness. Impact: My labels improved hallucination rate by 15% on test set and led to less defensive responses when refusing harmful requests.

I worked on a large scale Reinforcement Learning from Human Feedback (RLHF) pipeline to fine-tune a generative AI assistant to be safer and follow instructions better as a Tier 2 annotator and auditor. This involved ranking preferences instead of binary ranking as well as "red-teaming". Major Accomplishments & Skills Showcased Multi-Turn Context Reviews: Completed reviews of conversations with more than 10 dialogue turns to check that the model remained consistent and didn't provide contradictory/unsafe responses over longer conversations. Adversarial Red-Teaming & Edge Case Labeling: Human-written edge-case prompts created to elicit hallucinations/bias/too-much refusal. Counterfactual responses were written to train the model why a particular response was harmful instead of just labeling it "bad". Granular Preference Ranking: Performed side-by-side preference labeling with granular written explanations. Prioritized tone, succinctness, and strict adherence to safety guidelines over objective correctness. Impact: My labels improved hallucination rate by 15% on test set and led to less defensive responses when refusing harmful requests.

2025 - 2026

Education

U

University of Benin

Bachelor of Engineering, Civil Engineering

Bachelor of Engineering
2021 - 2025

Work History

A

Advanced Systems Global Technology

Frontend Developer

London
2025 - Present
I

Infodata

Frontend Developer

ogun
2025 - Present