An assistant that shows you where the answer came from.
Document and data assistants grounded in your own records. Every answer cites its source. When the answer is not in your documents, it says so instead of inventing one.
The problem with most of these
AI pilots fail in a consistent pattern. The assistant sounds confident, gets something wrong in front of a customer or a regulator, and nobody trusts it again. Gartner reported in January 2026 that at least half of generative-AI projects were abandoned after proof of concept by the end of 2025. Scepticism is now the default position, and it is earned.
The failure is almost never the model. It is that the system was built to always produce an answer, and an answer it cannot support is indistinguishable from one it can.
How we build them
Retrieval first, generation second. The assistant finds the relevant passages in your own documents before it writes anything, and every claim in the output is anchored to a specific source you can open and read.
When retrieval comes back empty or thin, the assistant reports the gap. It does not fill it. That is the single design decision that separates a tool your team will still be using in six months from one that gets quietly abandoned.
What it is good for
Finding the answer across contracts, policies, manuals and correspondence that nobody has time to read. Reconciling records that arrive in different formats from different places. Drafting a response that has to be consistent with what you have already said in writing.
What ships with every build
- Every claim in the output linked to the source passage it came from.
- An evaluation set built from your own real questions, run before handover, with the results published rather than described.
- A written record of what the assistant does not cover.
- Repository, hosting and API accounts in your name.
From $5,500
Every engagement starts with a paid scoping phase, because the honest answer to "can AI do this" sometimes turns out to be no, and you should find that out for a small amount of money rather than a large one.
What we do not build
Customer-facing support chatbots. The evidence on those is poor, the failure is public when it comes, and we would rather turn down the work than sell you a rollback.
Bring the documents your answers have to come from.
Kordal Systems is a design and engineering studio building websites, web applications and AI-integrated products for e-commerce brands and B2B companies.