JKD CODING
WhatsApp

Be Water, My Friend

Knowledge is water — it takes shape when it is contained

THE KNOWLEDGE DISTILLERY

JKD Coding provides knowledge extraction for large enterprises and SMEs: we distil the experience scattered across meetings, emails and the minds of veteran colleagues into executable, reusable rules that your AI employees can act on — the moment the meeting ends, the rules are on the wall.

“Absorb what is useful, discard what is useless.”

Absorb what is useful, discard what is useless — Bruce Lee defined knowledge extraction decades ago.

Four ways organisational knowledge evaporates

The veteran retires

Twenty years of judgement gone with a farewell party — the experience never left his head.

The minutes graveyard

Stored, but never read again. The four-thousandth file sits deep in a shared drive.

Fossilised SOPs

Outdated the moment they are written — documents trail reality, and new hires learn the wrong way.

The hollow AI

Without your knowledge, an AI is just a polite chatbot.

THE DISTILLERY · FEATURE

The journey of a single drop

Up top, the raw noise that falls every day; the single gold line in the middle is the hand that intercepts it; below, the rules that stay and do the work.

meeting-notes_v7_final(2).docxThe 38th cross-department meeting"He said we'd discuss it later…"Shared drive (4,200 files)The veteran retires next month217 emails back and forth
RULE · MEETINGSEvery meeting ends with "who + does what + by when" — missing any one, it reconvenes.
RULE · EXTERNALAny commitment is confirmed internally before it goes out — no one speaks for the company off the cuff.
RULE · DOCUMENTSOperating procedures are reviewed every 90 days — the writer signs, the reviser leaves a trace.
RULE · HANDOVERA new hire reads the rule wall within three days, then starts work.

Sample rules are generic illustrations, not any client's actual rule library

“It's not the daily increase but daily decrease — hack away at the unessential.”

Not daily increase but daily decrease — distillation is the art of cutting. A rule wall is carved, not hoarded.

The four-step distillation loop

01

Intercept

INTERCEPT

The day a meeting ends, its record goes straight into the still — before it evaporates.

Output: raw knowledge captured

02

Distil

DISTIL

AI and consultant cut together: AI refines, the consultant strips away the unessential.

Output: candidate rules

03

Crystallise

CRYSTALLISE

Human review puts rules on the wall — each one: context + action + owner.

Output: the rule wall

04

Reflow

REFLOW

AI employees work by the rules; monthly reviews sharpen them with use.

Output: organisational memory

“Using no way as way — having no limitation as limitation.”

Using no way as way — rules are distilled from real practice, not copied from templates.

PROOF OF PRACTICE

We don't teach distillation — we distil every day

252

working rules continuously distilled in our own governance system

Weekly

harvest-and-review rhythm, never skipped

6 languages

site-wide knowledge-governance experience, directly transferable

Who this is for

NGOs

Donor relationships and volunteer experience stop leaving with every staff rotation.

Social enterprises

Mission and operating experience carried through growth without deforming.

SMEs

Turn the owner's judgement into an institution the whole company can use.

Large enterprises

Cross-department knowledge governance: each team's tacit experience becomes executable rules for AI employees.

FAQ

How long does distillation take?+

The first round targets same-day delivery: the day a meeting ends, its rules are on the wall. The full loop — intercept, distil, crystallise, reflow — advances weekly by organisation size, then keeps cycling through monthly reviews.

Will our knowledge leak?+

No. The rules we distil belong to you and live inside your jurisdiction; we handle them only during the engagement with your authorisation, and step away after delivery.

How is this different from AI knowledge-management tools?+

Tools store; distillation cuts and shapes. Without the "who + does what + by when" rule structure, more data is still unexecutable — we deliver judgement, not a database.

How do distilled rules connect to AI employees?+

The rule wall is what an AI employee works from: the AI project manager acts by the rules and confirms internally by the rules, and monthly reviews feed new experience back as new rules — a closed loop.

Does the approach differ between large enterprises and SMEs?+

SMEs focus on institutionalising the owner's judgement; large enterprises need cross-department governance — each team's tacit experience unified into non-conflicting, executable rule sets.

How do we start?+

Start with a 30-minute diagnosis: inventory your organisation's most valuable — and most at-risk (evaporating) — knowledge assets, then decide the scope and rhythm of distillation.

START

From a drop of noise to a team of AI employees

STEP 1

Diagnosis

A 30-minute one-on-one to inventory your organisation's most valuable — and most at-risk (evaporating) — knowledge assets.

STEP 2 · THIS PAGE

Knowledge distillation

The four-step loop puts rules on the wall — knowledge finally has a shape.

STEP 3

Managed AI employees

AI project managers on site, working from your rule wall — we sell employees, not software.

Sample rules are generic illustrations, not any client's actual rule libraryJKD Coding · Using no way as way