Anthony Ogunbameru

A B.Sc degree holder in Computer Science, with a Master’s Degree in view in Cybersecurity.

Computer

Science graduate and AI coding specialist with strong hands-on experience in Python, code review, AI training workflows, benchmark task design, and cybersecurity. Proven background in evaluating, correcting, and improving AI-generated code through model evaluation, repository enhancement, production-readiness review, and AfterQuery benchmark projects across Project Silver, Project Pluto, Artifacts, Brick, Images, and Blueprint workflows. Skilled in Linux system administration, network security, web application security, SOC operations, incident monitoring, phishing analysis, and cybersecurity awareness.

Experience with tools and platforms such as Pentera, Forcepoint, Swimlane, Microsoft Defender for Endpoint, KnowBe4, Docker, GitHub, and Linux. Strong at writing clear technical explanations, identifying code quality issues, designing test-based validation, and applying software engineering best practices to improve correctness, structure, maintainability, and security. Comfortable working independently in remote, deadline-driven environments, with familiarity in Docker-based development, Linux tools, and modern AI-assisted development workflows.

Passionate about helping AI systems and software teams produce secure, reliable, and high-quality solutions.

Information Security Intern at Nigerian National Petroleum Company Limited (2024-04 – 2024-09)

• Gained in-depth knowledge and hands-on experience in cybersecurity/information security within the oil and gas sector during a six-month internship.
• Gained knowledge and experience in Linux system administration, network security, automated penetration testing, and web application security using tools like Pentera, Forcepoint, Swimlane, Microsoft Defender for Endpoint, and KnowBe4.
• Contributed to Security Operations Center (SOC) activities, including real-time monitoring of security events, incident response, and case management using SIEM and SOAR platforms.
• Conducted research and delivered presentations on cybersecurity topics, such as zero trust security and brute force attack prevention, enhancing communication and analytical skills.
• Participated in cybersecurity awareness training, phishing prevention exercises, and data center operations, gaining insights into industry trends and operational challenges.
• Strengthened problem-solving, time management, and teamwork skills through collaborative projects and practical tasks in a high-stakes energy sector environment.

Data Labelling, Video and Image Annotator at Self employed (2025-12 – 2026-02)

I worked as a data, video, and image annotator where I carefully labeled and reviewed different types of multimedia content to help train and improve AI models. My role required strong attention to detail, consistency, and the ability to follow detailed guidelines while identifying edge cases. I made sure the data was accurate and well-structured, contributing directly to the quality and performance of machine learning systems.

AI Trainer at Revelo (2026-02 – Present)

• Project Name: Behavioural Data (Full Stack Apps)

Designed and developed complex prompt strategies to guide AI models toward generating structured full-stack application solutions.

• Conducted multi-turn model interactions, refined requirements, evaluated UI, architecture, code correctness, accessibility, and documentation quality, and produced comparison reports on model behavior.
• Project Name: PR Writer

Reviewed AI-generated code for correctness, readability, maintainability, modularity, documentation, test coverage, and production readiness.

• Delivered constructive, pull-request-style feedback with practical engineering recommendations to improve implementation quality and model training outcomes.

AI Benchmark, Repository, Artifact & Evaluation Contributor at AfterQuery (2026 – Present)

• Project Name: Project Silver

Developed and improved production-style synthetic Python repositories for AI coding evaluation, with focus on realistic package structure, public entry points, service layers, test coverage, and non-boilerplate implementation quality.

• Authored and repaired benchmark task submissions by preparing instruction files, pytest suites, verifier configuration, Docker-based execution flows, oracle patches, and repository safety checks for review-ready submissions.
• Debugged failed validation runs involving repository-root detection, malformed patches, import errors, missing task metadata, oracle/NOP failures, and difficulty-check feedback, then revised the files to improve correctness and acceptance quality.
• Project Name: Project Pluto

Created terminal-benchmark style coding tasks with clear problem statements, deterministic tests, reward parsing, solution explanations, and validation instructions designed to assess software engineering reasoning in command-line environments.

• Reviewed model pass/fail trajectories, analyzed why agents solved or failed tasks, and adjusted edge cases, difficulty level, test coverage, and instructions to balance solvability, originality, and benchmark usefulness.
• Worked across multiple task domains such as reconciliation workflows, scheduling logic, configuration parsing, incident handling, returns processing, and code-repair challenges while maintaining strict verifier and template requirements.
• Project Name: Artefacts

Created structured deliverables for AI evaluation workflows, including task documents, worked solutions, project notes, review-ready files, and formatted outputs that followed detailed acceptance and quality guidelines.

• Organized technical content into clear sections, strengthened reasoning explanations, improved wording for reviewer clarity, and ensured outputs were consistent with project instructions and submission standards.
• Project Name: Brick

Supported coding and evaluation workflows involving modular task construction, repository inspection, bug analysis, and implementation review, with attention to maintainability, file structure, and reproducible validation.

• Prepared copy-paste-ready fixed files, reviewed failing logs, identified specific causes of failed checks, and converted feedback into practical changes that improved task reliability and engineering quality.
• Project Name: Images

Reviewed image and diagram suitability for task submissions, checking whether visuals matched the prompt, communicated the intended reasoning clearly, and avoided ambiguity or irrelevant visual information.

• Created and refined image-guide instructions for visual reasoning tasks, including requirements for English-language diagrams, clear labels, appropriate screenshots, and project-aligned supporting visuals.
• Project Name: Blueprint

Built task blueprints that translated broad project requirements into complete benchmark-ready specifications, including artifact type selection, reasoning skills, worked-solution structure, and quality-control checks.

• Expanded worked solutions into step-by-step explanations with sufficient detail, stronger justification, and clearer reasoning flow so reviewers and model evaluators could understand the expected solution path.

Bachelor of Science in Computer Science – Wesley University, Ondo West, Ondo, Nigeria (2021 – 2025)