Snr Code Reviewer - Node.JS

About OpenTrain

OpenTrain is a central job board for AI-training and data-labeling roles. We aggregate openings from many AI companies and labeling platforms so contributors can discover this kind of work in one place.

Creating an OpenTrain account is free and applying takes only a few minutes.

• OpenTrain connects skilled reviewers with AI projects that need accurate, reliable human evaluation.
• You will work as a contractor through OpenTrain's platform to review and improve annotations that train models.

About AI training and this work

AI models learn from human-prepared examples; your reviews directly shape how those systems behave. Tasks include annotating, testing, and evaluating model outputs so they’re accurate, safe, and useful.

This role focuses on code-response annotation: validating AI-generated Node.js code, confirming it follows prompts, runs correctly, and meets best-practice standards.

• Work is fully remote and typically flexible in schedule.
• Many projects require no formal certification, but this one requires deep Node.js experience and hands-on validation skills.

The role

You will audit annotator reviews of AI-generated Node.js code by building and running each snippet inside sandboxed containers or VMs. Your job is to confirm prompt compliance, correct execution, and adherence to security, performance, and testing standards.

When you find mis-ratings you will correct them and provide concise, constructive feedback so every review aligns with our quality rubric and improves the dataset that trains next-generation AI models.

• Position type: Remote contractor, part-time.
• Time commitment: 20+ hours per week.
• Pay: $24 USD per hour.

What you'll do day to day

Execute AI-generated Node.js snippets in a sandboxed environment, reproduce results, and verify they satisfy the original prompt and tests. Assess code for correctness, reliability, and maintainability.

Correct annotator scores when they are wrong and write short, actionable feedback to align reviewers with the rubric.

• Build and run code in containerized sandboxes (Docker or equivalent sandboxed VMs).
• Confirm prompt compliance, run unit/integration tests, and validate expected outputs.
• Audit for security issues (OWASP concerns, dependency CVEs, supply-chain hygiene) and performance pitfalls.
• Check that recommended tooling and runtime expectations are met: Node.js v18–v22, ES2023+ syntax, ESM/CJS, ts-node.
• Use structured rubrics and checklists to score reviews and file tickets when necessary.

Requirements

You must meet the following mandatory qualifications; we cannot accept substitutes or omissions.

This role demands advanced, hands-on Node.js and tooling knowledge plus practical sandboxing and testing experience.

• Experience: 7+ years in Node.js back-end or full-stack development, QA, or code-review roles.
• Node/runtime: Deep knowledge of Node.js internals (event loop, libuv, worker threads) and experience with Node.js v18–v22.
• Language/tooling: Modern JavaScript/TypeScript (ES2023+), ESM and CJS module systems, npm/pnpm, ts-node.
• Testing/debugging: Advanced use of Jest, Vitest, Mocha/Chai, supertest, coverage tools (nyc, Istanbul), and profilers (clinic.js, node-inspect).
• Security: Ability to detect injection flaws, insecure deserialization, dependency CVEs, and privilege-escalation issues; familiarity with npm-audit or Snyk.
• Performance: Experience with async patterns, streaming, clustering, worker threads, and caching strategies (Redis, in-memory LRU).
• DevOps/toolchain: Comfortable with Docker, containerized builds, CI/CD (GitHub Actions or GitLab CI), and observability stacks (winston/pino, OpenTelemetry).
• Proof-of-work validation: Able to spin up containers or sandboxed VMs to execute code and verify prompt compliance.
• Structured QA: Experience with rubric-based scoring, checklists, and ticketing/annotation tools such as Jira or Asana.
• Communication: Strong written English (B2+ CEFR) and a track record of clear, constructive feedback and mentoring.

Nice-to-have skills

The following are not required but will help you succeed and may give you an edge on complex review tasks.

• Exposure to LLM evaluation, RLHF pipelines, or experience with data-labeling platforms like Labelbox or Scale AI.
• Prior experience reviewing code as part of ML model evaluation or working directly on annotation quality improvement.

Who should apply and how it works

Apply if you enjoy hands-on validation, can reproduce and debug code quickly in isolated environments, and can provide concise, actionable feedback to reviewers.

As a contractor, you'll be assigned batches of annotated code reviews to audit. Each task requires running code, scoring the annotation according to a rubric, and providing corrective comments where needed.

• Worldwide applicants accepted; work is fully remote.
• Employment: Contractor, part-time.
• How to start: Create an OpenTrain account (free), apply to this listing, and complete any short qualification checks or sample reviews required by the project.