Companies Offering Remote Evaluator Jobs in 2026: Current Roles and Requirements

Quick answer: As of September 2026, eight companies and platforms have verifiable, current pathways into remote evaluator work: Welo Data (Welocalize), TELUS International AI Community, Appen, RWS TrainAI, OneForma (Centific), DataAnnotation, Outlier AI, and Alignerr. Below, each one is marked with what we could actually confirm: whether the company exists, whether it offers evaluator-type work, whether applications are currently open, and whether a matching vacancy was visible at the time of writing. Where a company publishes real pay or process detail on its own site, we quote it specifically instead of giving you a generic summary.

An earlier version of this article promised 21 companies without listing any of them. That was a mistake. This version lists fewer companies, but every one of them was checked against its own careers page, job board, or company documentation rather than copied from other “best of” lists.

Published: October 24, 2025  |  Last updated: September 24, 2026

Table of Contents

• How We Verified These Companies
• At a Glance: Comparison Table
• Not All “Evaluator” Work Is the Same
• Companies With Verified Evaluator Roles
• Two Companies That No Longer Operate Separately
• Platforms Worth Knowing, With Caveats
• How the Application Process Actually Works
• Before You Apply: Checklist
• FAQ

How We Verified These Companies

“Hiring” gets used loosely in this niche, so we checked four separate things for each company rather than treating them as one claim: whether the company or platform actually exists under the name shown, whether it offers evaluator-type work specifically (not just data labeling or transcription in general), whether applications are open right now, and whether we could find a specific, dated posting rather than just a general “join our community” page. Very few companies clear all four bars at once, because most run continuous, project-based signups rather than traditional job postings that open and close. We say exactly which bars each company clears below instead of calling all of them “hiring now.”

At a Glance: Comparison Table

“Not publicly confirmed” means we couldn’t verify that specific field against a primary source this session, not that the answer is no. Full detail and sourcing for every row is in the sections below.

Not All “Evaluator” Work Is the Same

“Remote evaluator” gets used as if it’s one job. It isn’t. The companies above actually span at least four distinct categories, and which one a company offers affects what the work looks like day to day: search quality rating (Welo Data, TELUS, Appen, RWS) means judging search engine results against published guidelines; AI response rating (DataAnnotation, Outlier AI, part of Alignerr’s work) means rating or comparing chatbot-style AI output instead of search results; content rating (OneForma’s Atlas Content Rater) means evaluating content and writing justifications for your decisions; and transcription and healthcare QA labeling (a large share of Alignerr’s current volume) isn’t evaluator work in the search-rating sense at all, even though it’s recruited through the same kind of platform. If you’re specifically after search quality rating, DataAnnotation and most of Alignerr’s current roles aren’t a match even though they’re evaluator-adjacent.

Companies With Verified Evaluator Roles

1. Welo Data (Welocalize)

Status: Current matching vacancy verified. Welo Data, the evaluator-recruitment arm of Welocalize, was running a live posting titled “Remote Internet Search Quality Rater – English (United States)” on its official Lever job board at the time of writing. The specifics that came directly from the posting, not from a summary: it’s classified as W2 part-time employment with biweekly pay, eligibility is limited to residents of about 18 named US states (the list ran roughly Alabama through Wisconsin, not nationwide), the start date was listed as ASAP, and applicants are required to sign an NDA and have a reliable computer, internet connection, and antivirus software. No prior search-rating experience was required, but the posting did require passing a client entrance exam first. A separate UK listing, “Scout Search Quality Rater,” was live at the same time under different terms.

2. TELUS International AI Community

Status: Evaluator role documented, applications open, specific listing not independently pulled this session. This isn’t a startup guessing at the work, TELUS International acquired Lionbridge’s entire AI and data-training division in a deal announced in November 2020, reportedly worth around $935 million and bringing over 750 employees onto TELUS’s payroll. That acquired team is what now runs TELUS’s AI Community program, and it’s the direct successor to what used to be marketed as “Lionbridge” search quality rater work (more on that below). Workers commonly refer to the search-rating program as “UEAT,” but we could not confirm that’s TELUS’s own official program name against TELUS’s own documentation this session, so treat it as an informal nickname rather than a verified brand. TELUS’s own qualification process is a real, multi-stage entrance exam rather than a quick quiz; our TELUS AI qualification exam guide walks through what that specific test covers.

3. Appen

Status: Evaluator role documented, applications open, specific listing not independently pulled this session. Appen runs search quality rater and AI evaluator projects across a large number of countries through its crowd platform. Its public careers page doesn’t list individual openings directly, instead it routes applicants through an “Explore open roles” gateway to a separate crowd-worker signup platform called crowdgen.com, where the actual project-level listings live. That two-step structure is why we could confirm the application channel is open but not pull one specific dated vacancy this session, the listings sit a layer deeper than Appen’s own public careers page. Before you get to that stage, our Appen qualification test guide covers what the entrance test actually asks.

4. RWS TrainAI

Status: Applications accepted on an ongoing basis, no fixed vacancy structure. RWS’s TrainAI Community recruits Online Raters, Data Annotators, Search Engine Evaluators, and Ad Evaluators through a standing community sign-up rather than individual postings that open and close. RWS’s own documentation lays out the actual evaluation steps: you submit background information (skills, interests, hobbies), then complete language proficiency testing to establish your language level, then take general machine-learning task tests that measure attention to detail, critical thinking, and performance on typical data-annotation work, not a generic aptitude quiz. RWS states explicitly that no fee is ever required to join, and questions about an application go to AICommunity@rws.com directly rather than a generic contact form. Because there’s no single “vacancy” with a close date, this counts as applications-open and evaluator-work-confirmed, not a specific current listing.

5. OneForma (Centific)

Status: Specific role verified, ongoing signup model. OneForma, operated by Centific, publishes named evaluator roles on its own site, including an “Atlas Content Rater” position where the actual work is rating content and writing justifications for your evaluation decisions in English, not just clicking a rating scale. Getting in requires creating an account with an email address, verifying it with a one-time code, and accepting OneForma’s platform agreements before you’re routed into the qualification flow for a specific project. That’s a continuous account-creation and qualification process, not a single posting with a close date.

6. DataAnnotation

Status: Evaluator-adjacent work confirmed, not search-specific. DataAnnotation runs a continuously open job board (published directly at dataannotation.tech/job-board/generalist) for rating and comparing AI-generated conversational responses, this is evaluator-type work, but it’s rating chatbot output quality, not search engine results specifically. Getting started involves a writing-quality assessment during onboarding rather than a search-rating exam. Contributors work as independent contractors and are paid weekly via PayPal once tasks are completed, not on a fixed salary schedule. If you’re specifically after search engine evaluation rather than AI response rating, treat this as a related but genuinely different category of work.

7. Outlier AI

Status: Applications open, evaluator-type work confirmed, some pay tiers officially published by country. Outlier AI, the contributor-facing platform associated with Scale AI, posts individual “opportunities” for response raters and agent-system evaluators, with pay and role type varying by country. This is one case where we found real, company-published numbers rather than worker estimates: Outlier’s own country-specific pages list pay directly, for example “Indonesian Voice AI Evaluator” at up to $7.50/hour and “Bengali Voice AI Evaluator” at $7.50/hour, published on Outlier’s own site for those specific roles. That’s not a blanket rate for the platform, pay is set per role and per country, but it means at least some of Outlier’s pay claims are company-sourced rather than scraped from a forum. Scale AI’s older Remotasks platform is being folded into Outlier, so if you land on a Remotasks listing elsewhere, check whether the same opportunity is now posted under Outlier before applying. Full opportunity listings require an authenticated account to browse, so we could not pull every current listing this session.

To put that $7.50/hour figure in perspective: at 15 hours a week, that’s 15 × $7.50 × 4.33 weeks ≈ $487/month. At 25 hours a week it’s roughly $812/month. This is an illustrative calculation based on one published rate for one specific role, not a guaranteed monthly income, actual hours depend entirely on project availability in your country and language, which Outlier doesn’t guarantee in writing.

8. Alignerr

Status: Company, hiring volume, and category-level pay verified directly; search-quality-specific role not confirmed this session. Alignerr, built on Labelbox’s data infrastructure, had roughly 5,600 open roles live on its own careers page at the time of writing, a real number pulled directly from the page, not an estimate. The roles we could actually see were concentrated in audio transcription and transcript editing (multiple languages: Hindi, Simplified Chinese, Spanish, German, Brazilian Portuguese, French, Italian) at a published $10-$35/hour, plus Healthcare AI Call Reviewer and QA Labeler roles for Central American Spanish speakers at a published $15-$20/hour. Those figures came directly from Alignerr’s own jobs page for those specific categories, not a third-party estimate. What we couldn’t confirm is a search-quality-rating-specific listing among what we pulled, Alignerr clearly does broader AI evaluation work, but don’t assume this is a drop-in substitute for the search-rater roles listed above without checking the current category list yourself.

Two Companies That No Longer Operate Separately

If you’ve seen these two names on other lists, including an earlier version of this one, here’s specifically why they don’t get their own entries above:

• Lionbridge: Lionbridge’s AI and data-training division, the part of the business that ran search quality rater work, was sold to TELUS International in a deal announced November 6, 2020, and reported at roughly $935 million. That unit had brought in around C$260 million in 2019, up 29% year over year, and employed more than 750 people at the time of the deal, which is why TELUS’s resulting AI Community program (listed above) is a substantial, established operation rather than a fresh start. Lionbridge itself continues to exist as a separate translation and localization company, but it is not a current employer for evaluator roles, and our own Lionbridge legitimacy review reaches the same conclusion independently, recommending Appen as the closest active alternative.
• RaterLabs: RaterLabs was a sister brand created by Leapforce, headquartered in Pleasanton, California, specifically to hire US-based search quality raters as part-time employees, with an expected average workload around 10 hours a week when tasks were available and a qualification test workers described as moderate difficulty. After Leapforce was acquired by Appen, RaterLabs was absorbed into Appen’s operations, and its former domain now resolves to a parked, for-sale GoDaddy page rather than an active company site. Our RaterLabs background and history page covers what the program actually involved, but Appen is where that work lives now.

Platforms Worth Knowing, With Caveats

These two come up constantly in search results for evaluator jobs, but they don’t fit the same category as the companies above without a caveat:

• Clickworker: Clickworker itself is a general microtask platform, not an evaluator-specific one. Search-evaluation-style tasks show up through UHRS, a separate Microsoft-run task layer that some Clickworker accounts get access to after building a track record on the base platform, not through a direct application or job posting for evaluator work. Treat “Clickworker for search evaluation” as a possible eventual path you earn access to, not something you can apply for directly on day one.
• Remotasks: As noted above, Scale AI has been migrating Remotasks contributors to Outlier. If you land on a Remotasks listing, check whether the same opportunity is now posted under Outlier before applying, since the older platform’s listings may no longer reflect current, active projects.

How the Application Process Actually Works

Timelines vary by company and by how much demand there is for your language and location at the moment you apply, so we’re not giving you a single number for approval speed. What’s consistent across the companies above, based on what each one actually publishes about its own process: you create an account and provide eligibility information (location, language, sometimes ID verification); you then take a qualification exam or entrance test specific to that company’s work, RWS’s process (background info, language proficiency, then task-specific ML tests) and Welo Data’s client entrance exam are two concrete examples of what that looks like in practice, not a generic aptitude quiz; approval and first-task timing then depend on current project demand in your region and language, which fluctuates and isn’t published as a fixed range by any company here; and pay structure varies by company, Welo Data’s listing above is W2 part-time employment, while OneForma, DataAnnotation, and most of the rest are independent contractor, project-based work. If a source, including an older version of this article, gives you an exact number of minutes for an exam or a guaranteed approval window, treat that as outdated or unverified rather than current.

Before You Apply: Checklist

• Confirm the company’s current careers page directly, not a third-party aggregator, before starting an application. For Appen specifically, that means going through crowdgen.com via Appen’s own site, not a lookalike domain.
• Check country and state/region eligibility before spending time on a qualification exam. Welo Data’s current listing, for example, is limited to specific US states, not the whole country.
• Note whether the role is W2 employment or independent contractor work, since this affects taxes and expectations. Only Welo Data’s listing above was explicitly W2; the rest are contractor arrangements.
• Don’t pay any fee to apply or to access a “starter kit.” RWS’s own TrainAI documentation states this explicitly, and it holds across this industry generally.
• Read our search quality rater exam sample questions guide if you’re preparing for a qualification test, since several of the companies above use a similar exam structure.

FAQ

Do all these companies require no prior experience?

Not universally. The Welo Data posting we reviewed explicitly states no prior rating experience is required. Others don’t make that claim either way in their public postings, so check the specific listing rather than assuming it applies across every company.

How long does the qualification exam take?

None of the companies above publish a fixed time on their official pages, though the structure differs meaningfully: RWS’s process is multiple stages (background info, then language testing, then task-based ML tests), while Welo Data’s is a single client entrance exam. Budget more time for a multi-stage process than a single test, and don’t rely on a specific minute count from a third-party source.

Can I get approved and start earning the same day?

We found no official source from any of these companies guaranteeing same-day approval or same-day payment. Payment schedules are set by each company: Welo Data’s listing specifies biweekly pay under W2 employment, while DataAnnotation pays weekly via PayPal once tasks are completed.

What’s the pass rate on the qualification exams?

None of the companies covered here publish an official pass rate. Figures circulating online for this aren’t sourced to the companies themselves, so we’re not repeating a specific percentage here.

Can I apply to more than one of these at the same time?

Generally yes, since most of these, OneForma, DataAnnotation, RWS, Outlier, and Alignerr among them, are independent contractor arrangements with different companies rather than exclusive employment. The exception is roles explicitly structured as W2 part-time employment, like the Welo Data listing above, where the employer’s own policies on outside work would apply.