Your company’s operational datais a business asset.

Lettus AI partners with organizations to improve frontier AI using real business workflows and institutional knowledge. Participating companies receive compensation aligned with the value of their contribution and may benefit from improvements connected to the use cases they help shape. Data is used to evaluate and refine AI capabilities in realistic enterprise environments.

We look for operationally mature companies with 10+ employees, established documentation practices, and high-quality operational data such as SOPs, knowledge bases, CRM records, project histories, and internal processes.

We focus on material that reflects how work is actually performed — standard operating procedures, process documentation, internal knowledge bases, workflows, templates, operational frameworks, and related business artifacts, along with field video, real-world audio, and images recorded where the work happens. The most valuable datasets capture real expertise, decision-making patterns, and repeatable workflows that help AI systems learn how companies operate in practice.

Yes. Alongside documents and structured records, we work with field and on-site video such as task capture, walkthroughs, and inspections; real-world audio such as calls, dispatch, and shop-floor or field recordings; and images including photos, scans, diagrams, and annotated captures. Media is sourced the same way as any other material: the contributing company owns it, scope and rights are agreed first, and it passes through privacy review before a prepared dataset is used.

Protecting partner data is a core principle of the program. Personally identifiable information is removed where relevant, access is restricted to authorized team members, and data is processed through isolated pipelines that prevent commingling across engagements. Media carries a different privacy surface — faces, voices, badges, screens, and location detail can appear incidentally in footage from a working environment — so it is reviewed for those specifically rather than assumed clear. Data is retained only for agreed periods, companies can review representative samples before use, and deletion requests are honored per the engagement terms. Personnel with data access use managed devices and follow established security controls and training requirements.

Companies retain full ownership of their underlying material, including original documents, video, audio, and images. A partnership grants defined, scoped permissions for an agreed use case rather than transferring ownership, and permitted use, retention, and deletion terms are written into the agreement before any dataset is prepared.

At application we ask what formats you hold, whether the company owns the material outright or it includes third-party content, whether identifiable people such as faces or voices appear in it, and whether consents or releases are in place. You do not need to have every answer resolved — "unsure" is a valid response and tells us what to review together. Material with mixed or unclear rights is worked through during scoping before anything is prepared.

Original material is what your company contributes — the source documents, footage, recordings, and images as they exist in your environment. The prepared dataset is what is produced from it inside a controlled processing environment: de-identified, scoped to the agreed use case, and reviewed by your team before use. Partnerships license the prepared dataset, and you review representative samples before it is used.

Shared material is processed according to the agreed use case and handling requirements. Depending on the dataset, PII may be removed and certain records may undergo synthetic rewrites that preserve workflow structure while reducing direct ties to original content; video, audio, and images go through a privacy review for identifiable people and incidental detail before a prepared dataset is used. Data is kept only for the agreed retention period, with opportunities to review samples before prepared data is used. Upon request or at the end of an engagement, data is deleted according to the agreed retention and deletion terms.

Compensation reflects the characteristics and value of the data contributed — including volume, workflow complexity, data quality, domain expertise, and how unique the operational knowledge is. Each partnership is evaluated individually, with payouts aligned to the overall relevance and signal strength of the dataset.

Lettus AI offers a referral program for introductions to qualified companies. If a referred company is selected and enters a partnership, the referrer may receive an award of up to US$50,000. To ensure attribution, submit the referral to data@lettus.ai before the company enters our evaluation process. Eligibility and award amounts depend on the opportunity and applicable program terms.

After a referral is submitted, the Lettus AI team reviews the company and guides them through an initial evaluation interview to assess fit for the program. Companies that meet qualification criteria move forward to partnership discussions, agreement finalization, and onboarding. Once onboarding is complete, they can begin participating.

Yes. We welcome referrals and partnership applications from companies anywhere in the world. Each opportunity is evaluated based on the quality and relevance of its operational data, partnership requirements, and active demand.

At this time, we primarily work with companies that operate in English and maintain documentation, workflows, and knowledge bases in English. Company representatives must also be able to communicate with the Lettus AI team in English throughout the partnership and onboarding process.