Brian M.

Key Skills

Software

Other

Scale AI

Top Subject Matter

Large Language Models (LLMs)

Rlhf Domain Expertise

safety and quality evaluation.

Top Data Types

Text

3D Sensor

Image

Top Task Types

RLHF

Segmentation

Freelancer Overview

Senior AI Trainer | Global AI Solutions (Remote). Core strengths include Other, Scale AI, and N. Education includes Bachelor of Science, Jomo Kenyatta University of Agriculture and Technology (JKUAT) (2021). AI-training focus includes data types such as Text and 3D Sensor and labeling workflows including RLHF, Segmentation, and Entity (NER).

Labeling Experience

Senior AI Trainer | Global AI Solutions (Remote)

OtherTextRLHF

Served as a Senior AI Trainer leading RLHF-focused data annotation and evaluation workflows to improve LLM reasoning and creative writing quality. Developed and refined annotation guidelines to reduce hallucinations and performed rigorous quality assurance on labeled outputs. Collaborated with machine learning engineers to identify edge cases and improve model safety through targeted red teaming. • Led a team of 15+ annotators on high-complexity RLHF projects. • Reduced model hallucination rates by 25% over six months via guideline refinement. • Maintained 99% accuracy on text and code generation QA checks. • Conducted red teaming with ML engineers to target safety gaps and edge cases.

2024 - Present

Data Labeling Specialist | TechAnnotate Corp

Scale AI3D SensorSegmentation

Performed large-scale data labeling for autonomous vehicle perception, including 2D/3D bounding box labeling and semantic segmentation for training datasets. Contributed to building a proprietary NLP dataset for Swahili and English to support regional language model improvement. Optimized labeling workflows using Python automation to increase contributor throughput while preserving label quality. • Labeled autonomous vehicle data with 2D/3D bounding boxes and semantic segmentation. • Helped develop an NLP dataset for Swahili and English language model training. • Increased individual throughput by 40% using Python-scripted labeling workflows. • Ranked in the top 5% for consistency and speed on Scale AI.

2022 - 2023

Junior Data Analyst | Insight Data Services

Text

Supported data preparation for chatbot optimization by cleaning and preprocessing datasets for sentiment analysis and market research. Assisted in creating training sets centered on intent recognition and entity extraction for downstream model learning. Contributed labeled training data readiness to improve conversational AI behavior and performance. • Cleaned and pre-processed datasets used for sentiment analysis and market research. • Assisted creation of chatbot training sets for intent recognition. • Supported entity extraction labeling for training/optimization. • Prepared structured datasets to enable model evaluation and optimization.

2021 - 2022

Education

Work History