What to know
About Sigma AI
Sigma AI is a Madrid-based multilingual AI data company founded in 2008 by Daniel Tapias and Nuria Gómez. It recruits contributors across many countries and languages for annotation, transcription, data collection and AI evaluation projects.
The company uses both corporate employment and flexible data-project arrangements, so read the employment type on the individual listing.
AI data training work at Sigma AI
Contributor work includes categorization, annotation, correction, transcription, conversation collection, voice recording, translation, localization and evaluation of AI-generated content. Many roles recruit native speakers of a specific language, including European, Asian, African and regional languages. Some assignments require professional or cultural expertise in addition to fluency.
A listing can be remote within one country rather than worldwide. The board may show dozens of language roles, but each has separate residence, equipment and availability requirements. Open the full description and check whether the work is a freelance project, temporary contract or employee position before applying.
What is it like to work at Sigma AI?
Remote data roles are typically project-based and may offer flexible scheduling. A campaign can nevertheless set weekly availability, productivity targets, fixed meetings or deadlines. Duration depends on the client and dataset, and a project can be extended, shortened or paused.
Being accepted for one role does not guarantee future Sigma assignments. Language demand changes, and there can be gaps. Public listings do not establish one universal rate or payment schedule. The written project terms confirm the amount, currency, billable unit, approval rules and payout timing.
Quality is central to transcription and language evaluation, so another linguist or a reference set may be used to review your work. The campaign rules explain whether corrections are allowed and how rejected items affect payment. Each campaign should be treated as a separate engagement rather than continuous access to the wider board.
Application process at Sigma AI
A Sigma AI application asks for contact information, location, a CV and the relevant language or professional background. Depending on the project, Sigma can use a language test, transcription sample, interview or guideline assessment. A role may also check your computer, headset or recording environment.
If you pass, you receive project-specific onboarding, which can include contractor documents, confidentiality terms, payment setup, training and calibration tasks. Passing a general language screen does not guarantee immediate production work if the campaign is not ready or its quota has filled. Written compensation, start-date and task-access details distinguish a confirmed Sigma project from an initial opportunity.
Sigma’s screening reflects the data type. A transcription project can test listening accuracy and spelling, while an AI-response role may ask you to compare answers against a rubric and explain the better result. Specialist assignments can also request education or professional evidence before a project test.
Sigma publishes both regular company positions and temporary contributor projects. The vacancy identifies the arrangement, required country and whether onboarding leads to a confirmed campaign or only to the contributor pool.
Sources: Company website
Information last updated August 2026.