Platforms we track

44 platforms are tracked here. 26 of them have open listings in our index. The rest keep a page because their status is worth knowing.

Live job feeds

• micro1Expert networkTalent pool351 live roles

  Public hourly rates on job posts, typically $50–$200/hr for experts

  Fast-growing Scale AI challenger ($500M valuation). Apply once, AI interview, then matched to projects.

  Join micro1 ↗details
• MercorExpert networkOpen roles250 live roles

  Public hourly rates on every role, typically $10–$120/hr (plus per-task projects up to $2K)

  One of the largest AI-work marketplaces; $10B valuation (Oct 2025). Rates and roles verified from their public listings.

  Join Mercor ↗details
• AfterQueryExpert networkOpen roles219 live roles

  Publishes a rate on the face of every listing on its public expert board, from $10–$30/hr on entry tracks to $200–$250/hr on specialist ones

  YC-backed expert data platform. Its public board showed 126 open listings on 2026-08-26, each printing its own rate, no account needed.

  Join AfterQuery ↗details
• xAI Specialist AI TutorsDirect AI-lab rolesOpen roles25 live roles

  Hourly ranges published on postings

  Direct contractor roles at an AI lab. Generalist hiring paused in 2025; specialist tutors (STEM, medicine, finance) still recruited.

  Join xAI Specialist AI Tutors ↗details
• AlignerrExpert network (Labelbox)Talent pool627 live roles

  States "$80/hr average pay" on its own homepage. Every listing carries an hourly band, 67 different ones, running from $10–$15/hr at the bottom to $200–$400/hr at the top

  Alignerr advertises each job many times over. Its board showed 5,530 listings on 2026-08-26 but only 469 separate jobs. Some are copied once per city, with the city name dropped into a single sentence of an otherwise identical posting, and every listing still says Remote. Others are simply repeated, word for word or lightly reworded. This directory lists each job once. Listings carry no date, and some workers report slow payment resolution, so apply with that in mind.

  Join Alignerr ↗details
• Meridial (Invisible)Freelance AI trainer projectsOpen roles769 live roles

  Wide bands (e.g. $6–$65/hr): exact rate set after evaluation

  Invisible Technologies' expert-contractor network. 800+ open freelance AI-trainer projects.

  Join Meridial (Invisible) ↗details
• TeracExpert panel & AI evalsOpen roles60 live roles

  Publishes a pay figure on every individual opportunity, from $2 to $400, though most of it is a one-off fee for a piece of work rather than an hourly rate. Its own landing pages advertise wider ranges by field, such as $60–$160/hr for finance, which no single listing matches

  AI-native expert panel blending market research with frontier-AI evaluation work. The public board at terac.com/opportunities is readable without an account. Its board showed 73 open opportunities on 2026-08-26. Only 17 of them pay by the hour, 46 pay a one-time fee and 10 pay per task.

  Join Terac ↗details
• Handshake AIExpert fellowshipOpen roles222 live roles

  Publishes an “up to” rate on all 203 of its roles and never a band, 24 different values from up to $25/hr to up to $300/hr, with up to $80/hr in the middle. These are ceilings, not typical pay

  PhD and graduate-focused AI training fellowship. Its own board showed 203 open roles on 2026-08-26, each on its own public page with a rate, no account needed.

  Join Handshake AI ↗details
• Babel AudioVoice data collectionOpen roles43 live roles

  Publishes a rate on every role, but the unit is an hour of audio delivered, not an hour worked. Conversation roles pay a flat rate per recorded hour of talking, which is close to an hourly wage. Transcription pays up to a stated figure per audio hour transcribed and voice acting a flat figure per recorded hour of finished audio, and both take several working hours to produce, so they are shown per task here and never as an hourly rate

  San Francisco voice-data vendor that records conversations, transcripts and voice acting in 20+ languages for AI labs. Its public board showed 43 open roles on 2026-09-05, each on its own page with a rate, no account needed. Payouts are weekly. Worker reviews put the conversation roles near the advertised figure once matching wait is counted.

  Join Babel Audio ↗details
• SME CareersAI training data and expert evaluationOpen roles735 live roles

  Checked 2026-10-04: every posting carries a structured rate. 1,694 of the 1,695 country postings were hourly in US dollars, and the other was a monthly staff salary in Bangladeshi taka on a hybrid role, which is not listed. The 735 postings listed here ran from 10 to 120 dollars an hour with a median of 30. SME's board prints most rates as a ceiling (Up to $50/hr), and this directory lists every one as a ceiling, including one role whose 1 dollar minimum SME's board prints as $1 - $70/hr. Rates are set per country. 22 of the 129 job titles listed here paid differently by country, and Materials Scientist ran from 30 dollars an hour in Spain, Italy, Portugal and Gibraltar to 95 in the United States, Germany and New Zealand. Three postings, at 2.50 and 3 dollars an hour, fall under this directory's 10 dollar floor, so they are not listed. SME's FAQ, read 2026-10-03, says rates are set per engagement

  Checked 2026-10-04: SME Careers describes itself in its postings as an AI data services company and subsidiary of SuperAnnotate, and on 2026-10-03 its site footer named SME Solutions Inc. Its public board at sme.careers/jobs showed 261 roles to a signed-out visitor, each with its rate, behind 1,695 country postings, and 1,598 of those postings were remote contract work. 955 of the 1,695 said in their own text that there was no immediate project for the role, and this directory does not list them. On 2026-10-03 a role's own panel listed one application step, an AI interview for that role, and applying sent a signed-out visitor to sign in or sign up first. Some postings say selection then adds a domain task and a recruiter interview

  Join SME Careers ↗details
• TuringAI training data and expert contractsOpen roles258 live roles

  Checked 2026-09-19: 30 of 226 postings stated a rate in their own text. Half were $15 or $20 an hour for AI quality analyst work in one language. Per-task rates ran from $30 to $400 on research, coding, voice and legal tasks. Three art roles estimated $7 an hour, under this directory's $10 floor, so they are not listed

  Checked 2026-09-19: San Francisco company that supplies coding, STEM and expert training data to frontier AI labs, according to its own help centre. Its public board at work.turing.com/jobs showed 226 roles, nearly all remote contract work. Its on-site and hybrid office roles are not carried here. Applying needs a Turing account

  Join Turing ↗details
• 10x TeamExpert networkOpen roles126 live roles

  Checked 2026-10-07: every AI trainer posting shows a euro hourly range, from 25 euros at the low end to 375 at the high end, for 8 to 20 hours a week of freelance work

  Checked 2026-10-07: 10x Team B.V., incorporated in the Netherlands, places senior European professionals as AI trainers for AI labs. Its terms say that on AI lab projects it contracts with the lab itself and engages the expert as an independent contractor. Its board also carries fractional executive roles, which are not listed here

  Join 10x Team ↗details
• DataAnnotationTask marketplace (Surge AI)Talent pool112 live roles

  Public recruitment pages advertise up to $25–$50/hr for generalists, up to $75–$150+/hr for software engineers, and role-specific specialist figures. These are advertised ceilings; assessment results and available projects determine actual work and pay

  Its public board has separate recruitment tracks with their own pay wording and application links. We refresh those pages in the jobs feed. Applicants complete assessments before project matching; these are ongoing talent-pool applications, not dated individual vacancies or guaranteed work.

  Join DataAnnotation ↗details
• Rise Data LabsAI training and annotationOpen roles88 live roles

  Checked 2026-09-12: published hourly bands on public job pages. Roles without a published rate are excluded here

  Checked 2026-09-12: public roles cover annotation, coding, languages and domain expertise. The board remains the source of availability, including older listings that are still advertised

  Join Rise Data Labs ↗details
• RWS TrainAIAI data communityOpen roles65 live roles

  Checked 2026-10-04: 45 of the 47 TrainAI postings state an hourly rate in their own text, and none fills Lever's salary field. Of the 38 postings outside the Search Engine Evaluator family and the talent pool, 37 quote US dollars, from 4 to 36.45 an hour, and the 8 Legal Annotators roles pay 13.50 to 36.45 US dollars an hour. The 8 Search Engine Evaluator postings give only a ceiling, up to 4 to 15 US dollars or up to 10 to 14 euros an hour. Of the 38, 7 pay under 10 US dollars an hour, so this directory's 10 dollar floor keeps them off the list

  Checked 2026-10-04: RWS's public Lever board carried 47 postings, all on its TrainAI team. Of those, 20 were posted in September 2026, the newest on 2026-09-29, among them 8 Legal Annotators postings, one per language or script, posted on 2026-09-16 with a stated duration to the end of October. Another 15 date from before June 2026, including the 8 Search Engine Evaluator roles posted on 2024-10-14 and a talent-pool posting from 2025-03-14, which this directory does not list. Lever files 46 as temporary or contract roles and 1 as a part-time employee role in the United States

  Join RWS TrainAI ↗details
• LionbridgeTranscription and data collectionOpen roles45 live roles

  Checked 2026-10-07: 54 of 56 postings carry a structured rate in US dollars. 48 are hourly, from 7 to 45 dollars, and 7 of those are under this directory's 10 dollar floor and are not listed. 6 are one-time payments of 8 to 20 dollars for smartphone surveys and data collection

  Checked 2026-10-07: Lionbridge's AI Data Training team posts freelance transcription, annotation and data collection projects on its own Lever board. Its site names Lionbridge Technologies, LLC

  Join Lionbridge ↗details
• ClaruAI training and data collectionOpen roles40 live roles

  Publishes rates on individual role pages. Accepted footage is paid per hour of media and shown per task here. Below-floor roles are omitted.

  Checked 2026-09-12: public roles span annotation, model evaluation and video capture. The open board determines availability because a still-listed role carried an expired validThrough date.

  Join Claru ↗details
• BraintrustAI work on a freelance marketplaceOpen roles30 live roles

  Checked 2026-09-12: this feed takes only AI-training roles with structured hourly or per-task budgets. Fixed-price projects and unpriced roles are excluded

  Checked 2026-09-12: Trustpilot reviewers describe long screening processes and unsolicited contact after applying. These are individual reports about the marketplace, not verified outcomes for every employer

  Join Braintrust ↗details
• TransPerfect DataForceAI data collectionOpen roles25 live roles

  Published payment is often per accepted recording, photo, story or study session. An hour of accepted media is a task unit, not an hourly wage.

  Checked 2026-09-13: public voice, video, photo and writing projects. A February 2026 Trustpilot reviewer reported unpaid voice recordings after declining further re-recording. DataForce replied offering to investigate the rejected recordings.

  Join TransPerfect DataForce ↗details
• InnodataAI training and evaluationOpen roles19 live roles

  Contributor roles publish hourly pay. Repeated state copies with the same title and compensation collapse to one role. Canadian rates remain in CAD.

  Checked 2026-09-13: the board mixes contributor projects and staff roles. Only the specified contributor titles and AI Solutions Engineer are included. Glassdoor's Generative AI Associate page showed 2.4 of 5 from 61 reviews, with individual reports of layoffs and flat pay.

  Join Innodata ↗details
• VettoExpert networkOpen roles14 live roles

  Checked 2026-10-07: most roles pay a fixed US dollar fee per approved task or deliverable, from 30 to 60 dollars, paid in local currency. The hourly figures beside the fees are Vetto's own estimates and are not shown as rates here

  Checked 2026-10-07: expert review and scenario writing projects in coding, health, law, finance, education, marketing and AI safety. Applications ask screening questions, and some postings add a short exercise

  Join Vetto ↗details
• GrayderLife science expertsOpen roles13 live roles

  Checked 2026-10-05: every public role page publishes a USD hourly band, from 10 dollars for translation review to 50 to 100 dollars for clinical annotation

  Checked 2026-10-05: clinical and research roles for doctors, scientists and regulatory specialists. Closed roles stay online marked closed and leave the board

  Join Grayder ↗details
• WorkadaAI training and annotationOpen roles7 live roles

  Checked 2026-10-07: every public role page publishes a USD hourly rate, from 17 dollars for data labeling to 100 dollars for finance and life science roles. Pay is sent three times a week through Stripe

  Checked 2026-10-07: six open roles in data labeling, video metadata, science charts, finance and life sciences. Every role includes a short interview with Workada's team

  Join Workada ↗details
• iMerit ScholarsAI training and evaluationOpen roles6 live roles

  The public feed publishes explicit hourly rates alongside low-paying country variants. Roles below the directory floor and amounts without a currency or unit are omitted.

  Checked 2026-09-12: Scholars posts AI evaluation and annotation contracts. A Trustpilot reviewer on 2025-12-22 reported roughly 2000 dollars unpaid for an OpenAI project. This is an individual worker report, not a verified finding about all contracts.

  Join iMerit Scholars ↗details
• Sovrano AIAI evaluation and data servicesOpen roles3 live roles

  Checked 2026-09-20: every published role carries a structured rate. 21 were hourly, at 60 US dollars an hour for the slide and deck expert roles, 12 euros an hour for the native-speaker evaluator roles, 8 to 12 euros an hour for the speech data roles and 5 to 15 US dollars an hour for robotics annotation. One paid 500 US dollars per approved task. Four paid 1,000 euros a month, stored as an annual salary the way this directory stores every monthly figure. Eleven of the hourly roles state in their own payment terms that the rate is for a pilot and that production rates are still to be confirmed, so the euro figures on those may change. The roles at 8 euros an hour, and the monthly ones, fall under this directory's 10 dollar floor, so they are not listed

  Checked 2026-09-20: Barcelona company, Sovrano AI, S.L., that runs an AI data services platform matching evaluators with client projects in annotation, RLHF, labelling and evaluation, according to its own terms of use. Its board at app.sovrano.ai/opportunities showed 36 records, 26 published and 10 closed. A signed-out visitor sees all 36 with their rates, and the 26 published ones with an Apply button, the closed ones reading View instead. Contracts are remote and independent, and applying needs a Sovrano account

  Join Sovrano AI ↗details
• Outlier (Scale AI)AI training marketplaceOpen roles1 live roles

  Public opportunities advertise up-to hourly ceilings. Rates depend on assessment, location and project; onboarding or non-core work can pay less

  Public project listings refresh automatically. General interest forms and listings with conflicting pay are excluded. Acceptance does not guarantee tasks or the advertised rate.

  Join Outlier (Scale AI) ↗details
• Fleet AIDomain expert AI evaluationOpen roles

  Fellow roles publish typical hourly ranges. Final compensation depends on domain, experience, location and engagement.

  Checked 2026-09-13: eight active domain expert Fellow contracts. Screening questions vary by role. The thin review footprint was not used for a sentiment rating.

  Join Fleet AI ↗details
• G2iAI coding evaluationOpen roles

  Explicit hourly ranges appear in the role descriptions. Weekly workload hours are kept separate from compensation.

  Checked 2026-09-13: public AI code-evaluation contracts require experienced engineers. Trustpilot showed 4.2 of 5 from 57 reviews. A Glassdoor reviewer in June 2024 reported reduced client rates. G2i replied that it acted as recruiter and the client set the terms.

  Join G2i ↗details
• Pasiflora AIAI evaluation and data projectsOpen roles

  Rates come from each open role's headline pay field. Approved-participant and recorded-video payments stay task units. Separate proposal fees remain in the role description.

  Checked 2026-09-13: three open roles and 124 talent-pool entries. Two open roles have a headline amount. The remaining role says its rate is set by the client project. No verified review page was used.

  Join Pasiflora AI ↗details

Tracked, but not currently listing

• EthosExpert networkTalent pool

  Publishes an exact rate and a weekly cap on each opportunity page, seen from $50/hr up to $225/hr depending on the role

  a16z-backed expert network (May 2026). Voice-agent onboarding; work is matched inbound rather than browsed.

  Join Ethos ↗details
• AuraOneRelay boardOpen rolesnot listed here

  AuraOne advertises 3,346 specialist roles under its own name, and its own public feed gives 1,034 of them a Mercor job id.

  Checked 2026-09-20: each of its 3,346 listings carries a sourceJobIds value in its own feed. 1,034 are Mercor list_ ids, 852 are AuraOne's own, and 1,460 are ids whose source could not be confirmed. All 30 Mercor ids sampled opened a Mercor role page, and it names AuraOne Human Data as the hiring organisation on all 3,346. Every listing is marked evergreen with no expiry and a status of active, and the newest posting date in the feed is July 2026.

  details
• CrowdGen by AppenCrowd projectsOpen rolesnot listed here

  CrowdGen by Appen was removed after our August 2026 review found missing pay information in the public project feed we checked and concerns in payment reports. Its current project terms need to be checked directly.

  Appen's contributor platform (ex Appen Connect). Modest volume, clean structured pay data.

  Join CrowdGen by Appen ↗details
• HumanSignalAI training and annotationTalent pool

  Checked 2026-10-04: its contributor page shows example rates for US residents, $22 to $45 an hour for AI response evaluation and $15 to $35 for data collection, both paid per accepted task, and $60 to $180 for domain expert review. It says each project shows its rate before work starts. Its Greenhouse postings give other figures in their text, from up to $6 an hour on several country data collection postings to up to $100 an hour for US candidates on its remote one.

  Checked 2026-10-04: HumanSignal, Inc., the company behind the open source labeling tool Label Studio, runs this contributor sign-up for AI response evaluation, data collection and domain expert review. You create an account, build a profile and pass an identity check, and some projects add a skill screen or credential review. Projects are shown only after sign-in. The page says payouts are weekly in one place and that the cadence varies in two others, and it names no payout method. The terms linked under its sign-up form are the customer terms for its software, and we found no contributor terms. This directory does not read its Greenhouse board, so nothing from it is listed here.

  Join HumanSignal ↗details
• LuelData collection marketplaceOpen roles

  Checked 2026-10-04: its contributor page showed six tasks with a rate each, $12 an hour for American English conversations among three people and $7.50 an hour for partner conversations in Uyghur, Tibetan, Hmong, Jamaican Creole and Guarani. Its terms say conversation pay is counted per participant by duration, that no rate is locked when you record or upload, and that the rate paid is the one in force when a submission is approved. Its contributor page says payouts are processed daily.

  Checked 2026-10-04: Luel Inc. runs a data marketplace that pays contributors to record conversations, video and photos and licenses them to AI companies as training data. It is in Y Combinator's Winter 2026 batch, and in May 2026 it announced a $31.2 million seed round led by General Catalyst and Lightspeed Venture Partners. Its terms ask that you be at least 18. You pick a task, record it in its app, and read a short phrase before each session so it can check your voice against your earlier recordings. We found no interview or skills test. Full task details sit behind sign-in and there is no public feed, so nothing from it is listed here.

  Join Luel ↗details
• MeredenExpert networkOpen roles

  Checked 2026-10-10: its site lists 33 roles, and 32 of them show an hourly band in US dollars, from $15 to $30 for robotics reviewers and driving experts up to $100 to $150 for olympiad medallists. Medical experts and STEM PhDs are shown at $80 to $140, legal experts at $70 to $130, and its general mathematics, coding and STEM evaluator roles at $20 to $40. The 33rd invites you to propose a role of your own and names no rate. Its role pages say the vetting task is paid and that payouts go out weekly. It publishes no dates on its roles, so no rate here is read from a dated listing.

  Checked 2026-10-10: Mereden describes itself as a marketplace that matches credential-verified experts with AI labs and research teams for paid evaluation and data work. Its role pages say you are engaged by Mereden as an independent contractor, and its terms say paid work comes under separate project terms given at onboarding. No page we read names the company behind it, an address or a governing law, and its pages sign off only as meredenai.com. The domain meredenai.com was registered on 2026-08-13, according to the RDAP registry. That date is the start of the domain's current registration, not the company's founding. The steps are an account, a profile with your resume and work authorization, then a short paid vetting task in your domain. This directory does not ingest its roles, so nothing from it is listed here.

  Join Mereden ↗details
• MindriftFreelance AI tasks (Toloka)Open rolesnot listed here

  Mindrift listings advertised $15 to $100 an hour, but those figures belong to a specialist board most contributors never reach, and the open work underneath it is not the kind of pay this directory exists to point people at.

  Toloka's freelance platform: all Toloka crowd tasks migrated here. Its listings are no longer carried here; see the notes above for why.

  Join Mindrift ↗details
• NobleStark HatchExpert networkTalent pool

  Checked 2026-10-04: its Hatch page lists five tracks with an hourly band each, $120 to $200 for agentic software engineering, $80 to $160 for code generation and review, $150 to $300 for cybersecurity and red-team work, $100 to $250 for math and formal reasoning, and $140 to $220 for custom eval and benchmark authoring. It says the trial task is paid but not how much, and that payouts are weekly. It publishes no individual openings, so no rate here is read from a listing.

  Checked 2026-10-04: Hatch is the contractor network of Noble Stark LLC, which sells datasets, benchmarks and expert evaluations to AI companies. Its legal pages give an address in Newark, Delaware, and Delaware governing law, and do not say where the company was formed. Its pages give two founding years. Its about and press pages say 2024. Its careers page says 2020, and describes a technology company building digital products and talent placement solutions, with 100+ employees in 6 locations. Its about and Hatch pages say its network has 5,000+ vetted specialists. The domain noblestark.com was registered on 2025-08-08, according to the RDAP registry. That date is the start of the domain's current registration, not the company's founding. The steps are a short form, a paid trial task of about 2 to 3 hours, a 30 minute review with a domain lead, then a contractor agreement and an NDA.

  Join NobleStark Hatch ↗details
• OneForma (Centific)Crowd projectsOpen roles

  Publishes no amount anywhere. States only that you see your rate before accepting a project, and that payment carries no minimum threshold

  Data collection, annotation and linguistics projects. Its board showed 28 open projects on 2026-08-26, browsable without an account, but not one of them states a rate, which is why nothing from it can be listed here.

  Join OneForma (Centific) ↗details
• OpenTrain AIRelay boardOpen rolesnot listed here

  OpenTrain AI carries about 1,300 open roles and publishes a rate on most of them, but when this directory counted on 2026-09-05, about a sixth of its titles were near-exact copies of roles live on four other platforms.

  Checked 2026-09-20: 1,306 open roles on its own board. On 2026-09-05, 238 of its roughly 1,350 titles were near-exact copies of roles live on Mercor, micro1, Alignerr and AfterQuery. Every page sampled names OpenTrain AI as the hiring and contracting organisation, and the apply link stays on app.opentrain.ai.

  details
• PeritVoice data collectionOpen roles

  Checked 2026-10-04: its recording page pays per approved audio hour, $20 for read sessions, $22 for prompted speech, $24 for everyday rooms, $28 for conversations and $20 to $32 for accents it is short on. An audio hour is recorded audio, not time worked, and its own example puts a 20 minute session at about $7. Its careers page shows $17 to $45 per audio hour on some transcription and alignment roles, and $210 to $365 a month on its India rater and annotation contracts, which its job data states as 20,000 to 35,000 Indian rupees.

  Checked 2026-10-04: Perit is the brand of Zyno AI Inc, a training-data company in Y Combinator's Fall 2026 batch, according to its own about page and YC. It pays people to record speech on their phones and runs a bench of transcribers, aligners and voice raters on its own platform, Foundry. Recording needs a sign-up and a mic check, and its page says there is no interview. Paid annotation work starts after you pass its training and test modules. Approved sessions are paid within seven days once your balance passes $10. Recording sessions are shown only after sign-in. Its careers page lists 11 roles, six with a structured monthly salary in Indian rupees and five with pay only in text or on application. This directory does not ingest them, so nothing from it is listed here.

  Join Perit ↗details
• Pila8Voice data collectionOpen rolesnot listed here

  Pila8's job pages advertise an estimated $10 to $50 an hour, but when we signed up on 2026-09-23, the only project its app offered us paid $6 an hour.

  Checked 2026-09-23: three open roles, Audio Contributor, Conversational Recording Contributor and Single-Person Recording Contributor. Each page carries JobPosting markup with no salary field and the posting date 2026-03-16. Projects are listed only inside the app, after sign-up and a short first test.

  Join Pila8 ↗details
• ProxionExpert networkTalent pool

  Checked 2026-10-04: its experts page and its one posting advertise $75 to $300 an hour, set by specialization, seniority and project complexity, and say pay can rise for reviewers and project leads. It says the technical test is paid and that contractors invoice and are paid bi-weekly. It publishes no individual openings with their own rate, so no rate here is read from a listing.

  Checked 2026-10-04: Proxion, Inc., a Delaware corporation headquartered in New York, hires finance professionals as independent contractors to build and review financial models, write expert questions and answers, and grade AI outputs. It asks for at least two years in investment banking, private equity or structured finance and at least 15 hours a week. The steps are an application, a screen, a paid technical test, an interview and onboarding. Its privacy policy names Persona for identity checks, Checkr or another provider for background screening where permitted, and Deel for payments. It has one generic posting on its own site and no job board, so nothing from it is listed here.

  Join Proxion ↗details
• TELUS Digital AIRater and linguist gigsOpen rolesnot listed here

  TELUS Digital AI had 67 open search and ads rating roles across 33 countries on 2026-09-20, but it does not publish what the work pays, so there is nothing here to copy.

  Checked 2026-09-20: TELUS Digital's AI community board. Roles are the familiar rater families, Online Data Analyst 26, Media Search Analyst 17, Quality Assurance Rater 11 and Personalized Internet Assessor 6. Its detail pages carry no JobPosting markup and no posting date, and the board answers 403 to scripts.

  Join TELUS Digital AI ↗details
• Welo Data (Welocalize)Rater & linguist gigsOpen rolesnot listed here

  Most of what Welo Data lists is search and ads quality rating rather than AI training, a third of its openings are more than six months old, and the rates that are published are the lowest of any feed this directory carried.

  Welocalize's AI-data arm. Steady flow of search/ads rater and AI data roles.

  Join Welo Data (Welocalize) ↗details