Hope A.

Key Skills

Software

Don't disclose

Top Subject Matter

Multimodal AI (video tracking) and RLHF for LLMs

Web3 security

smart contract parsing

Top Data Types

Video

Image

Text

Document

Top Task Types

Tracking

Fine-tuning

Segmentation

Text Generation

Text Summarization

Transcription

Computer Programming/Coding

Object Detection

Freelancer Overview

Technical Data Annotator & AI Trainer (Freelance / Distributed Remote). Brings 3+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Custom Annotation Pipelines and Don't disclose. Education includes High School Certificate, Mid-Town Comprehensive Secondary School (2012) and Certificate, The5school (2022). AI-training focus includes data types such as Video, Computer Code, and Programming and labeling workflows including Tracking and Fine-tuning.

Labeling Experience

Technical Data Annotator & AI Trainer (Freelance / Distributed Remote)

VideoTracking

Conducted high-fidelity multimodal video data annotation for edge AI vision models, including hyper-precise atomic hand-to-object spatial movements and physical interaction tracking. Performed RLHF-oriented evaluation and rewriting of multi-turn prompts and LLM responses to align outputs with factuality and safety constraints. Executed systematic backend code evaluation and logic tracing to tag bugs, classify execution paths, and support code-generation model training. • Multimodal video labeling and physical interaction tracking. • RLHF prompt + response writing (evaluation/scoring and rewrites). • Code evaluation and execution-path logic tracing (Node.js/Rust). • Quality compliance scoring on sandboxed calibration sets (98%+).

2024 - Present

Advanced Smart Contract & Protocol Logic Labeling (Independent Datasets)

Don't discloseFine-tuning

Curated structured training rows for specialized smart contract parsers across Sui (Move), Solana (Anchor), and Starknet (Cairo). Labeled vulnerability patterns, logical deadlocks, and gas optimization profiles to build clean fine-tuning datasets for blockchain engineering AI assistants. Systematically documented the dataset so it could be reused for training and validation workflows. • Dataset curation for smart contract parsers (Sui/Move, Solana/Anchor, Starknet/Cairo). • Vulnerability and logical-deadlock labeling. • Gas optimization profile labeling. • Fine-tuning dataset documentation and preparation.

Present

Education

Work History