Kyle L.

Key Skills

Software

Mercor

Data Annotation Tech

Top Subject Matter

Financial Compliance & Risk Analysis

AI Model Evaluation

Tax Domain Expertise

Top Data Types

Text

Document

Image

Top Task Types

RLHF

Fine-tuning

Evaluation/Rating

Prompt + Response Writing (SFT)

Freelancer Overview

Team Lead - AI Training Finance Expert. Brings 8+ years of professional experience across complex professional workflows, research, and quality-focused execution. AI-training focus includes data types such as Text and labeling workflows including Evaluation, Rating, and RLHF. Education includes Bachelor of Business Administration, University of Georgia, Terry College of Business (2019).

Labeling Experience

AI Training Tax Expert

MercorText

As an AI Training Tax Expert with Mercor, I verified LLM outputs related to complex tax and compliance issues for global mobility. I evaluated nuanced financial reasoning in model responses to ensure technical and regulatory accuracy. I generated synthetic datasets and benchmarks to support AI understanding of specialized tax scenarios. • Authenticated responses on international taxation matters and residency reporting. • Created high-quality "Golden Answers" for critical test cases. • Identified subtle errors in tax logic and model compliance assessments. • Enhanced the reliability of AI-generated solutions for tax professionals.

2026 - Present

Team Lead - AI Training Finance Expert

MercorText

As a Team Lead - AI Training Finance Expert at Mercor, I applied finance expertise to assess potential harm and financial impact of model outputs in high-stakes domains. I reviewed and provided feedback to subject matter experts, facilitating team training and workflow optimization. I developed and evaluated complex financial test cases to challenge AI model reasoning in professional-grade tasks. • Oversaw multi-disciplinary teams to ensure project adherence and technical accuracy. • Coordinated deliverable hand-offs between multiple managers for consistent output. • Led strategic development of stress tests targeting model robustness. • Focused on precise quality control for training data in finance-related LLMs.

2026 - Present

AI Prompt and Response Evaluator

Data Annotation TechTextRLHF

As an AI Prompt and Response Evaluator at Data Annotation Tech, I conducted complex RLHF to optimize model accuracy and instruction-following. I performed multi-turn red-teaming and adversarial evaluation to probe for ethical flaws, bias, and edge-case vulnerabilities. I designed targeted prompts and adversarial tests to stress model safety and identify potential failures. • Implemented human-in-the-loop feedback for continuous model improvement. • Authored adversarial prompts for red-teaming and pressure-testing safety filters. • Identified edge cases related to fairness, bias, and constraint boundaries. • Helped design guidelines for robust AI evaluation in safety-critical tasks.

2024 - 2025

Education

Work History