Ezra L.

Key Skills

Software

Argilla

Data Annotation Tech

Dataloop

CVAT

Label Studio

Top Subject Matter

LLM Output Evaluation, RLHF Preference Ranking & Structured Data Annotation

AI Agent Development & Tool Calling Systems

Text Data Quality Assessment & Ambiguity Resolution

Top Data Types

Text

Image

Computer Code Programming

Top Task Types

RLHF

Classification

Entity (NER) Classification

Freelancer Overview

Kairos — AI-Powered Intelligent Calendar Web App (LLM extraction, ambiguity resolution, and evaluation feedback collecti. Brings 2+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include DeepSeek API. Education includes Bachelor of Engineering, Guangdong University of Technology (2022). AI-training focus includes data types such as Text and labeling workflows including RLHF.

Labeling Experience

Kairos — AI-Powered Intelligent Calendar Web App (LLM extraction, ambiguity resolution, and evaluation feedback collection)

TextRLHF

Developed an AI-integrated pipeline to extract structured calendar event data from free-form text, using LLM JSON mode and tool calling with regex fallback and confidence-scored outputs. Designed an ambiguity resolver to collect explicit accept/modify/reject feedback and support an inter-annotator-style review flow. Implemented and documented failure modes such as hallucinated dates, over-refusal on ambiguous input, and confidence miscalibration to inform preference-ranking and evaluation practices. • Ingested user text and produced structured event fields with confidence scoring • Collected user feedback (accept/modify/reject) per extracted event • Logged model/user interaction metadata to PostgreSQL for quality analysis • Built an agent dialogue system with tool definitions and SSE streaming for step-wise reasoning visualization

2024 - 2025

Education

Work History