Available Jobs
Available Jobs on Neevo can vary quite a bit. Broadly speaking, you'll be asked to label, upload, or validate text, audio, or images/videos.
More specifically, that means you may be asked to record yourself reading text, categorize audio based on key characteristics or sentiments, draw bounding boxes around objects in images, or identify key people, and places in passages of text (among other tasks).
Below are a few examples of the kinds of work you may be asked to do on Neevo. However, it's important to note that these examples only represent a small fraction of available job types on Neevo.
Text-Audio Validation (TAVA)
In this type of job, your task is to listen to the provided audio input and compare it with the corresponding text to ensure accuracy. Although this template can accommodate tasks such as multiple-choice questions, it is most commonly used for binary (yes/no) verification, where you confirm whether the audio matches the text or identify any discrepancies. Precision is crucial in this role, as the data helps improve speech recognition systems and ensures high-quality, reliable results.
Mean Opinion Score Test (MOST/MOS)
In this type of job, your role is to provide a subjective evaluation of audio data. You will assess various aspects such as clarity, naturalness, pronunciation, emotional expression, and overall audio quality. The Mean Opinion Score is a standardized method used to rate the perceived quality of speech or sound on a numerical scale, typically ranging from poor to excellent. Your feedback plays a vital role in refining audio systems, training AI models, and improving voice synthesis technologies. Attention to detail and consistency in your evaluations are essential for producing reliable and actionable results.
Pronunciation Validation (PRVA)
In this type of job, your task is to evaluate whether all words in the given phrases are pronounced correctly. This involves checking for proper articulation, accurate pronunciation, and adherence to the intended language or accent standards. Attention to detail is crucial to ensure the audio meets quality and consistency requirements.
Scripted Monologue Data Collection (SCMDC)
In this type of job, your primary task is to record yourself reading the provided text aloud. The goal is to collect high-quality audio data that aligns with the given script, ensuring clear pronunciation, consistent pacing, and natural intonation. This data is often used to train AI models for speech recognition, natural language processing, or other audio-based applications. Precision and clarity in your recordings are essential to maintain the integrity and usability of the collected data. This type of job might require installing the Neevo app on your smartphone, as some of these tasks cannot be performed on a computer.
ABX
In an ABX job, you compare two or three audio or text samples (A, B, and X) to determine which of the first two (A or B) matches the third (X) or to identify which of the two is preferred—or if neither is preferred. This process helps evaluate subtle differences or similarities in pronunciation, audio quality, or linguistic features. By completing each step methodically, you contribute to identifying issues, refining workflows, and ensuring the data meets high-quality standards for further analysis or application.
CSAT
This task might involve evaluating speech or written data in a Customer Satisfaction (CSAT) survey, where you assess the clarity, tone, and appropriateness of responses in relation to customer interactions.
Image/Video Collection and Validation
In this type of job, you must gather or review multimedia content to ensure it meets specific quality standards, relevance, or labeling requirements. The adaptability of Generic Jobs makes them essential for collecting precise and diverse data to meet specific project objectives.
