How Nigerian Artisans Can Earn From First-Person Video of Their Trade

Imagine a tailor in Aba, a mechanic at Ladipo Market, or a cook in a busy buka, all earning extra income simply by doing what they do every day. This is now possible through the creation of first-person video data, a crucial component for training the next generation of artificial intelligence and robotics. If you are a Nigerian artisan with a trade you are skilled in, this article explains how you can participate and earn from this emerging field.

Is recording videos for AI legit for Nigerian artisans?

Yes, recording videos of your trade for AI training is a legitimate way for Nigerian artisans to earn. Companies like TrustAm pay individuals for approved recordings of their everyday work. This process involves capturing first-person video and audio of your tasks, which is then used to train AI systems. It's important to work with reputable platforms that prioritize your safety, privacy, and fair payment. TrustAm, for instance, has a clear process for verification and payment, ensuring that contributors are compensated for their work. You can learn more about the legitimacy of this work in Nigeria here.

The demand for real-world data is growing rapidly as AI developers strive to create systems that can understand and perform tasks in complex, unpredictable environments. First-person, or egocentric, video data is particularly valuable because it captures the world as a human experiences it, including actions, interactions, and the nuances of different trades. This type of data is essential for training robots to perform tasks in factories, assist in homes, or even for AI that can provide guidance to workers. You can find more details on egocentric video data collection in Africa.

AI and robotics are not just for advanced labs; they need to understand the skills and practices of everyday workers. By recording your trade, you are contributing to the development of AI that can be more practical, helpful, and accessible. This data helps AI learn skills such as intricate repairs, precise cooking techniques, or detailed craftwork, making AI tools more useful for everyone.

What kind of work can be recorded?

Almost any hands-on trade or everyday work activity can be recorded for AI training. TrustAm currently focuses on trades like tailoring, auto mechanics, phone repair, cooking, hair braiding, welding, carpentry, and market trading in Lagos. The key is that the work should be performed by a person, and the recording should capture the task from their perspective. This means the video shows what the artisan sees and does while performing their job.

Whether you are a phone repairer in Computer Village, a hair braider in a salon, or a market trader managing your stock, your daily activities provide valuable insights. Even common tasks like preparing ingredients for a meal, welding metal, or cutting fabric can be part of a dataset. The goal is to collect diverse examples of how people perform tasks in real-world Nigerian settings. This diversity is crucial for building AI systems that are robust and can function across different conditions and cultural contexts. For example, tailors, mechanics, and cooks are among the trades that can provide essential data.

The data collected includes not only the visual aspect of the work but also the associated speech. If you are explaining a process, conversing with a customer, or even just narrating your actions, this audio is also valuable. Nigerian languages like Pidgin, Yoruba, Igbo, and Hausa, alongside English, are all in demand. This helps AI systems understand and respond in local languages, making them more accessible to a wider population. The Nigerian speech data collected is vital for this purpose.

How does the recording process work?

Participating is straightforward. You will need a smartphone that meets certain requirements (Android 9 or newer, 1080p camera, 8GB free space) and will be provided with a head strap to hold your phone securely while you work. The process is designed to be minimally intrusive, allowing you to perform your trade naturally. Before you start recording, you will read a consent line on video in your own language, ensuring you understand and agree to the terms.

During recording, a random 3-digit code is read aloud at the start, and both hands are shown to ensure correct setup. The phone's motion sensors (IMU) and GPS data are recorded alongside the video. TrustAm uses automatic checks to filter out poor-quality recordings, such as black frames, frozen video, or instances where the phone was not in use. After the automatic checks, a trained reviewer examines each minute of footage. You are paid only for the minutes that pass this quality review. If a minute is cut, you receive a reason and can appeal the decision. This multi-stage review ensures the data is high-quality and useful for AI developers. You can find more details on how people, places, and minutes are verified on our website.

The data collected is a rich combination of egocentric video, audio, and motion data. This holistic approach provides AI and robotics teams with a comprehensive understanding of tasks. For instance, understanding how a mechanic manipulates tools requires seeing the actions, hearing the sounds of the engine, and tracking the movements of the hands. This is similar to the data collected in large-scale projects like Ego4D, which aims to capture human activities from a first-person perspective.

What are the safety and privacy measures?

TrustAm takes the safety and privacy of its contributors very seriously. All contributors undergo verification using their BVN or NIN, matched against national records, along with a live selfie. Payments are made directly to the contributor's bank account or TrustAm account in their name, with a one-person-one-account policy. This ensures that payments go to the correct individuals and helps prevent fraud.

Workplaces are also registered by TrustAm field agents, and consent is obtained from the business owner before any recordings begin. Each registered workplace has a shop-front photo and a map pin with a defined boundary. For recordings involving conversations, all speakers must give their consent before the recording starts. Bystanders' faces are automatically blurred in the recordings to protect their privacy. Raw recordings are stored securely for only 90 days for quality assurance and appeals, after which they are deleted unless a specific contract requires longer retention. Reviewers access media with blurred faces and masked identities, and sensitive personal information like BVN/NIN is stored as a salted hash plus the last four digits. This commitment to data protection aligns with Nigerian data privacy regulations, such as those overseen by the Nigeria Data Protection Commission (NDPC).

Contributors have full control over their data. They provide explicit consent before their first task, and the specific consent version and date are included in the buyer's manifest. Anyone can request to stop future licensing of their recordings or ask for their data to be deleted at any time. This ensures that contributors remain in control of how their work is used. The NDP Act GAID 2025 also outlines the principles for data processing in Nigeria, which TrustAm adheres to.

How are contributors paid?

Contributors in the Lagos pilot are paid ₦5,500 per approved hour of recording. Payment is only for the minutes that pass quality review and is disbursed in naira to the contributor's own bank account with no fees charged by TrustAm. The process is designed to be transparent and fair, rewarding quality contributions. Work is opened in batches as buyer orders come in, meaning availability can vary.

It is important to understand that this is not a guaranteed full-time salary, but rather an opportunity to earn extra income by contributing valuable data. The pay is for approved minutes, and not all recorded minutes may pass the quality checks. This aligns with the nature of data collection for AI, where quality and relevance are paramount. TrustAm never asks for payment to join, for equipment, or for activation. Potential contributors can learn more about how to earn by recording their work at TrustAm's earn page.