Samuel O.

Key Skills

Software

Anno-Mage

Top Subject Matter

LLM behavior evaluation

Safety Domain Expertise

and fine-tuning dataset preparation

Top Data Types

Document

Image

Video

Top Task Types

RLHF

Prompt + Response Writing (SFT)

Text Generation

Freelancer Overview

Reinforcement Learning (RLHF) / Human-in-the-loop (HITL) Senior AI Scenario Designer, Neural Systems Lab. Brings 5+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Weights & Biases, Python-based evaluation scripts, and N. Education includes Master of Science, Imperial College London (2024) and Bachelor of Science, University of Edinburgh (2023). AI-training focus includes data types such as Text and labeling workflows including RLHF, Prompt + Response Writing (SFT), and Text Generation.

Labeling Experience

Reinforcement Learning (RLHF) / Human-in-the-loop (HITL) Senior AI Scenario Designer, Neural Systems Lab

TextRLHF

Designed and architected 500+ edge-case scenarios to evaluate LLM reasoning quality and support RLHF-style improvement loops. Implemented Human-in-the-Loop and reinforcement-learning oriented evaluation practices to reduce hallucinations and stress-test AI safety boundaries. Led dataset creation efforts to enable fine-tuning for vertical assistants in regulated domains. • Architected edge-case scenario sets for LLM reasoning evaluation. • Developed and executed Red Teaming protocols to surface conversational AI security issues. • Built multi-turn dialogue trees to guide HITL evaluation of assistant behavior. • Co-led fine-tuning dataset development for FinTech and Healthcare verticals using diverse data collections.

2024 - Present

AI Content & Interaction Designer, Vanguard Automation

TextPrompt + Response Writing (SFT)

Created interaction personas and style guides that function as training/evaluation targets for customer-facing AI assistants. Authored prompt engineering iteration materials and automated Python-based evaluation pipelines to validate generation quality across assistant releases. Managed cross-functional workflows to ensure prompt and behavioral guidelines were translated into deployed agent behavior. • Designed AI interaction personas to define expected assistant conduct. • Implemented automated testing pipelines for prompt engineering iterations. • Authored style guides for model personality, tone, and brand alignment. • Coordinated UX and ML teams for smooth agent deployment readiness.

2022 - 2024

Technical Narrative Designer, Arcade Dynamics

TextText Generation

Developed branching logic for non-player characters using early-stage generative text models, producing structured dialogue variations for downstream use. Optimized NLP scripts to generate real-time responses in interactive environments. Worked on narrative logic that effectively guides model output selection among multiple scripted paths. • Built branching NPC dialogue logic using generative text models. • Optimized NLP response-generation scripts for immersive, real-time interaction. • Produced structured text variations aligned with narrative/game flow requirements. • Supported iterative refinement of generated responses for coherent agent behavior.

2022 - 2023

Education

Work History