Frontend Engineers: $65-$190/hr from Frontier AI Labs

Open application· 102 spots this round

$65-$190/hr frontend engineering work, on your schedule

Review AI-generated UI code, components, and frontend architecture the way you'd review a PR before merge. Catch the accessibility violation, the re-render storm, the state bug that only shows in production. The judgment that ships interfaces that actually feel right is exactly what AI labs need.

Fully remoteYour scheduleWeekly pay

Apply nowApply once, get matched on a rolling basis. No prior AI experience needed.

Trusted by top research companies

Rated by experts worldwide

Hi, we're Zac and Jack, the founders of Terac. We want to talk to you directly, because you are the most important part of what we're building.

Terac is a community of experts. People who have spent years getting good at something specific and hard. The world is about to need more of you, not less. As AI takes on more of the world's work, the bottleneck shifts to the people who actually know what they're talking about.

Expert labor is the rarest resource in the world right now, and it is shockingly hard to find. The companies that need a frontend engineer's eye on a component that breaks accessibility spend weeks chasing people, paying placement fees, and settling for whoever is available. Meanwhile thousands of qualified people are sitting with knowledge that no one ever asks for.

That gap is what we're here to close. Every project that lands on Terac is routed to the people who actually know the answer, on their schedule, paid fairly, and only when the work is verified. No middleman taking a cut of your time. No vague gigs. No chasing checks.

We care about every single person in this community. If you join Terac, you're not a row in a database to us. We read the feedback. We answer the emails. We will fight for you when a customer is being unreasonable, and we will be honest with you when something on our side is broken. The quality of this panel is our entire company, and we owe you a serious bar.

If you've made it this far, here is what we're asking: claim your profile. Put your expertise on the record. Let the world's most ambitious teams come find you for the work only you can do.

Zac & Jack

Founders

Frontend Engineering questions

Still curious? Write to us at support@terac.com.

Niche specializations are often the most valuable, because frontier models struggle most with deep, non-tutorial-level knowledge. If you work day-to-day with WebGL shaders, WASM module boundaries, or the Web Animations API, you are exactly the kind of expert who can catch plausible-but-wrong model output that a generalist reviewer would miss. Demand varies week to week, so you may see fewer tasks than a React generalist would, but the tasks you do see will be matched to your specialty.

The tasks are calibrated to senior-level judgment, not syntax recall. You will be asked to evaluate things like architectural tradeoffs in component design, accessibility tree correctness, performance profiling interpretations, and whether a model's reasoning about browser rendering pipelines holds up under real constraints. Junior-level syntax checks are a small fraction of the work; the higher your seniority, the more of the task mix will be relevant to you.

You will review and annotate real code artifacts: JSX and TSX components, CSS-in-JS and plain CSS, Webpack or Vite configuration files, accessibility audits against WCAG 2.2 criteria, and written explanations of how the browser's event loop or paint pipeline behaves. You will also produce worked examples that demonstrate expert reasoning, such as walking through why a specific React reconciliation pattern causes unnecessary re-renders, with annotated code showing the fix.

Both of those credentials are recognized and strengthen your profile during onboarding verification. The CWAS is particularly relevant for tasks involving WCAG compliance review and ARIA implementation. Google Developer Expert status signals domain depth in specific ecosystems like Angular or Web Performance, and that context is used to route tasks to you that match those areas.

No. All tasks use synthetic or sanitized examples constructed specifically for training purposes, and you are never given access to any real company's codebase, repository, or proprietary design system. The concern about inadvertently reviewing confidential employer work does not apply here because the artifacts are purpose-built for evaluation, not sourced from real products.

Why your expertise matters

Frontend engineers make daily judgment calls that are nearly impossible to encode in a spec: when a component API is too clever for its own good, whether a CSS architecture will survive a design system migration, or whether a given React pattern will cause subtle hydration bugs at scale. Frontier models generating UI code frequently produce output that passes surface-level review but fails on production concerns like bundle size, accessibility trees, or runtime performance on mid-range Android devices. Terac needs engineers who have shipped real products to catch those failures before they propagate into model weights.

How pay works

Pay toward the top of the $65-$190/hr band reflects demonstrated specialization - engineers with deep experience in performance profiling, accessibility compliance (WCAG 2.2 / ARIA), design system architecture, or a specific framework ecosystem like React Native or Next.js App Router tend to command higher rates because the evaluation tasks are more complex. All work is fully remote, compensation is hourly, and payment is released on verified task completion rather than on a fixed schedule.

What the work looks like

A sample of the frontend engineering work you would pick up. Every project is scoped, remote, and paid on verified completion.

Review a model-generated React component for prop API design flaws, unintentional re-render triggers, and missing error boundaries, then annotate each issue with a short rationale a junior dev could learn from.
Evaluate a code-generation model's output for a responsive data table against real-world constraints: keyboard navigation, screen reader announcement order, and overflow behavior on 320px viewports.
Write a canonical worked example of a custom hook that manages async state with proper loading, error, and stale-data states, structured so a model can learn the idiomatic pattern rather than a naive implementation.
Rate five model-generated Tailwind component snippets on a rubric covering semantic HTML, color contrast ratios, and whether utility classes will conflict with a typical design token override layer.
Identify failure modes in AI-suggested Webpack configuration advice, specifically cases where the suggested code-splitting strategy would break dynamic imports under a Turbopack migration.
Create a set of adversarial prompts designed to probe whether a code-generation model understands the difference between controlled and uncontrolled form inputs in React, including edge cases around defaultValue hydration.