Vildan A. Navis · Embedded operator for maritime yards
Corporate AI Brain
Institutional memory for maritime yards — knowledge that stays when people leave · working model
Working model
The quiet leak — knowledge walking out the door
What a build actually cost vs. the quote, which subcontractor overran which block, why a bid was won or lost — today that lives in 3–4 people's heads. When an estimator or project manager leaves, it leaves with them, and the next quote starts from zero. Nothing in the yard captures it; the brain does — and makes it answerable in a sentence.
Ask the yard's memory
0
knowledge domains captured
0
to recall · any question
0
retained when people leave
0
to stand up · vs. lost today
Ask the yard's memory tap a question — what the brain returns once fed your record →
The eight memory domains tap any domain or node for the full read →
The thesis · why this is the asset to protect
Capability stays in the yard. Knowledge stays in people — until it walks.
Your yard can already build its flagship product — that engineering capability is institutional, it lives in the yard. But the commercial and cost intelligence around it — real build cost vs. quote, subcontractor track records, estimating assumptions, why each bid landed or didn't — is tacit, held by a handful of people. A family yard's sharpest risk isn't a machine breaking; it's an estimator retiring. The brain turns that tacit knowledge into a private, queryable institutional memory that compounds with every build instead of resetting with every departure.
From tacit to recallable · how it works
01 · Capture
Feed what already exists
Build files, post-mortems, estimating sheets, subcontractor records, bid outcomes — plus a public-record seed. A light habit per build keeps it current.
02 · Embed
Private, structured memory
Documents become a searchable knowledge layer (vector + structured), on the yard's own controlled infra. Role-based access — commercial sees pricing/bids, PMs see build history.
03 · Recall
Ask in plain language
"What did the last chem tanker cost vs. quote?" "Which subcontractor delayed block assembly?" "Why did we lose the 12k bid?" — answered in a sentence, with the source.
04 · Compound
Every build makes it smarter
Each project adds memory; quotes get sharper, subcontractor choices safer, the win-rate read clearer. Knowledge stops leaving — it accrues.
The brain remembers. The embedded operator builds and runs it. A document store isn't an asset — the discipline that captures each build, structures it, and makes it answerable for the next quote is. The model proves the capability; standing it up and embedding the habit is the engagement.
Why this is an AI/IT capability — not a shipbuilding pitch
A traditional yard has no system that remembers like this — knowledge lives in people and spreadsheets, and resets when they leave. This is data/AI plumbing: ingest the yard's own records, embed them privately, expose recall over them — run by an AI operator, not a naval architect. Adopted in days. Fed the yard's own record — public seed plus internal build data — the brain answers cost-vs-quote, subcontractor and margin questions no spreadsheet can — and keeps the institutional knowledge inside the yard for good.