Timotej F.

Key Skills

Software

Don't disclose

Top Subject Matter

Software Engineering - Python Code Evaluation & AI Training Data

Cybersecurity - Secure Code Review, Red Teaming & Vulnerability Analysis

AI Safety & LLM Evaluation - Prompt/Response Review, RLHF & Technical Annotation

Top Data Types

Computer Code Programming

Top Task Types

Computer Programming/Coding

Freelancer Overview

I have hands-on experience contributing to AI training data through Cipher/Shipd, where I create hard coding challenges designed to test model reasoning, security awareness, and software-engineering accuracy. My work includes building realistic backend and cybersecurity-focused tasks, writing clear problem specifications, designing validation checks, and iterating on submissions through review. Recent challenge topics include webhook authenticity in billing event intake, delegated approval isolation in case handoffs, and report asset isolation in project exports. What sets me apart is my combination of software engineering, cybersecurity, and AI-evaluation skills. As a Cyber Engineering student with strong interest in purple-team security, I am comfortable analyzing edge cases, identifying subtle vulnerabilities, and creating precise technical content for model training. I bring strong attention to detail, structured reasoning, and practical coding ability, which help me produce high-quality training data that improves how AI systems understand, generate, and evaluate secure code.

Labeling Experience

AI Coding Challenge Creator – Cipher/Shipd

Computer Code ProgrammingEvaluation/Rating

I contribute to Cipher/Shipd by creating high-quality Python coding challenges used for AI training and evaluation. The work involves designing realistic software-engineering and cybersecurity-focused tasks, writing clear challenge specifications, building test cases and validation checks, and reviewing whether solutions satisfy functional, security, and edge-case requirements. My contributions focus on hard Python challenges involving backend security, access-control isolation, webhook authenticity, and secure data handling. I ensure each task is deterministic, technically accurate, and difficult enough to test advanced model reasoning. Quality measures include clear acceptance criteria, reproducible checks, edge-case coverage, code-quality review, and revision based on reviewer feedback.

2026 - Present

Education

Work History