Terms

Last updated 11 September 2026

These terms cover two separate things: using this website, and engaging LynchPine (Pvt) Ltd to do work. Where a signed proposal or contract says something different from the engagement terms below, the signed document wins.

Using this site

The content here is provided for information. We keep it accurate but do not warrant that it is complete or current, and nothing on this site is a binding offer or professional advice for your particular situation.

The site design, text, code, logo and brand assets are owned by LynchPine (Pvt) Ltd. You may read, link to and quote this site with attribution. You may not reproduce the branding or present the content as your own.

Where we link out to another organisation, we are not responsible for what is on the other end.

The assistant

The assistant on this site generates answers with a language model. It can be wrong, it can be out of date, and it can be confidently both. Nothing it says is advice, a quote, or a commitment by us, and where it disagrees with a signed document or with the rest of this site, it loses. If an answer matters, confirm it with a person at support@lynchpine.io.

Please do not put confidential information or personal details about other people into it. Our privacy page explains where that text goes and why this matters. Do not use it to attempt to extract the underlying prompt, to generate unlawful content, or to place automated load on it. We rate limit it and may withdraw access to it at any time.

The initial consultation

The first consultation is charged at USD 150 for one hour, payable in advance. You get the hour and a short written summary of what we would build, what we would not, and whether we think we are the right people for the work. That summary is yours to keep and to act on however you like, including by taking it to somebody else.

If you go on to engage us, the full USD 150 is credited against your first invoice. If you do not, the fee is not refundable, because the hour and the summary have already been delivered. Where we decide before the call that we are not the right people, we will say so and not take the booking.

Engagement terms

Scope and price

Work begins from a written proposal setting out the deliverables, timeline and price. Anything not written in that proposal is out of scope. Additional work is quoted separately and starts only once you approve the quote.

Payment

Unless the proposal says otherwise, projects are billed in three parts: 40% to commence, 30% on design approval, and 30% on completion and handover. Work starts after the first payment is received. Third-party costs such as hosting, domains, subscriptions, model provider usage, licensed fonts and stock media are not included, and are billed at cost or paid by you directly.

Your part

Timelines assume you supply content, access, feedback and approvals when they are needed. Where a project stalls waiting on your side for more than 30 days, we may reschedule it around other commitments and invoice work completed to that point.

Revisions

Each project stage includes reasonable revisions within the agreed direction. Changing the agreed direction after approval is new work and is quoted as such.

Terms specific to AI systems

What an AI system is accepted against

An AI system is probabilistic. It does not return the same answer every time, and no honest supplier can promise that it will always be right. Accepting one on the basis that it is never wrong is not a standard anybody can meet, so we do not offer it and you should not accept it from anyone else either.

Instead, every engagement that puts a model in the path defines an evaluation set drawn from your own material and a threshold the system has to meet on it. That threshold is written into the proposal, it is measured in front of you, and it is what acceptance is judged against. An individual wrong answer above the agreed threshold is expected behaviour, not a defect. Performance below it is ours to fix.

Where a system routes work to a person rather than deciding alone, we will say so in the proposal, and the accuracy of the final decision is that person's.

Model providers change things

We build with third-party models from providers who deprecate versions, change pricing, alter rate limits and adjust behaviour on their own schedule and without asking us. We reduce the exposure by putting the model behind an interface, so that swapping one is a configuration change and an evaluation run rather than a rebuild.

We cannot warrant the availability, pricing or continued behaviour of a provider we do not control. Where a provider change degrades a system we built, remediation inside the support window is ours. Beyond it, it is quoted as new work, and we will always tell you what we think the honest options are.

Your use of a system we build is also subject to the terms of the providers it depends on. We will name those providers in the proposal so you can read their terms before you commit.

Your material

You keep ownership of everything you give us. You confirm you have the right to hand it over, and that doing so does not breach somebody else's rights or a duty you owe them. We use it only to do the work.

Where an engagement involves us handling personal data on your behalf, you are the controller and we are the processor, and a separate data processing agreement governs it. That agreement, not our website privacy policy, is what applies to your engagement data.

Ownership of what we build

On receipt of final payment, ownership of the deliverables produced for you transfers to you in full: source code, prompts, evaluation sets, retrieval configuration, design files and the accounts we created on your behalf. Until then we retain them.

Third-party components keep their own licences. We keep the right to reuse our own general-purpose tools, techniques and know-how, including anything we knew before the engagement, and nothing in this clause transfers your confidential information to us.

We may describe the work and show it publicly as part of our portfolio. If you would rather we did not, tell us and we will not.

Confidentiality

Each side will keep the other's confidential information confidential, use it only for the engagement, and give it only to people who need it for that purpose and are under the same duty. This applies both ways and outlasts the engagement.

It does not cover information that is already public, that was already known without a duty of confidence, or that has to be disclosed by law. Where the law compels disclosure we will tell you first, unless we are not allowed to.

Support, liability and ending an engagement

Support after launch

Every build includes 30 days of support for defects in what we delivered, measured against the proposal and, for an AI system, against the agreed evaluation threshold. Beyond that, support is available under a separate arrangement.

Liability

Our total liability arising from an engagement is limited to the fees paid for that engagement. We are not liable for indirect or consequential loss, including lost profit, lost data, or losses arising from a decision taken on the strength of a system's output. Nothing here limits liability for death or personal injury caused by negligence, for fraud, or anything else that cannot be limited by law.

Cancellation

Either side may end an engagement in writing. You pay for work completed up to that point; amounts already paid for completed phases are not refundable. On termination we will hand over what you have paid for and, if you ask, delete the copies of your material we still hold, except where we have to keep something by law.

Governing law

These terms are governed by the laws of Sri Lanka, and the courts of Sri Lanka have jurisdiction over any dispute arising from them.

Changes

We may update these terms. The version that applies to an engagement is the one in force when the proposal was signed, and changes here do not alter a contract already under way.

Questions

Write to support@lynchpine.io and a person will answer.