Innodata

How it works

Innodata is a Nasdaq-listed data engineering company (ticker INOD) with more than 36 years in the data business, and today it sells training data, model evaluation and annotation to large technology companies building generative AI. Innodata AI jobs for individuals are recruited through what it calls its Subject Matter Expert (SME) on Demand programme: part-time, remote, project-specific work that Innodata describes plainly as not a traditional full-time role.

The roles are posted on Innodata's careers site and its public Greenhouse job board. Most carry titles such as Generative AI Associate, AI Evaluator, AI Training Specialist, Data AI Annotator or Content Evaluator, usually tagged "Flexible Hours", and most are posted separately for an individual US state. The duties listed are the standard spread of AI training work: rating model outputs against evaluation questions, ranking and labelling content, writing prompts and responses, rewriting and summarising text, transcription, translation and data gathering.

The eligibility bar on the generalist postings is higher than the pay suggests: a bachelor's degree or higher in a humanities specialisation is required, a master's or PhD is strongly preferred, and English at professional or expert level (C1/C2). Specialist and bilingual evaluator roles appear alongside them at higher rates, each with its own rate stated on the posting. If you are weighing it against other AI training work that suits writers and educators, read the rate on the specific posting first.

Worth knowing

Good:

• Innodata is listed on Nasdaq and traces its business back more than 36 years, which puts it at the more established end of a market full of young startups
• Nearly every posting on its board states the pay, so you know the rate before you apply
• Innodata publishes a job-scam warning on its careers site, says it never asks for payment, banking details or sensitive personal information during the application process, and gives a dedicated address (verifyjoboffer@innodata.com) for checking an offer. Our guide to spotting fake AI training jobs explains why that matters in this sector
• Indeed reviewers of the Generative AI Associate role name the flexible start time and working from home as the main upsides

Less good:

• Almost every flexible-hours posting on Innodata's board at the time of checking is located in a US state; applicants elsewhere will find little to apply for
• The rate on the generalist postings is low for a role that requires a humanities degree and prefers a postgraduate one
• Generative AI Associate reviewers on Indeed describe high productivity targets, close monitoring and monotonous work, and some report being let go from projects without warning
• There is little independent contractor discussion of Innodata on Reddit compared with the gig-style platforms, so most outside evidence comes from job review sites

About this page

Facts above come from Innodata's homepage and careers site (innodata.com and innodatacareers.com) and the job descriptions on its public Greenhouse board, checked 29 September 2026. Innodata publishes a rate on each posting but no platform-wide figure, and the postings do not state payment method, payout cadence or contractor classification, so those fields are left empty rather than guessed.

Worker reports come from Indeed reviews and are attributed in the sentence that makes each claim. We have no referral arrangement with Innodata.