Technology / Knowledge Architecture

An organisation that cannot represent what it knows cannot reason about what it does.

A folder is not a memory. A vector DB is not an organisation. We design the layer in between.

Companies don't have a data problem. They have an epistemic infrastructure problem.

Decisions, expert reasoning, contradictions, abandoned hypotheses, version drift, the things people know but never wrote down — none of it lives in a CRM, a wiki, or a vector DB. AI sits on top of that debt and amplifies it.

Knowledge architecture is the layer that decides how organisational information becomes computable: how it is typed, sourced, versioned, decayed, contradicted, retrieved. Without it, AI is fluent and ungrounded.

There is no off-the-shelf pipeline.

The right shape depends on what knowledge the organisation actually produces, who creates it, where decisions live, and what the AI underneath is meant to do. Two companies in the same industry rarely need the same pipeline.

KVA runs the engagement: we map the knowledge surface, design the pipeline, structure the resulting OBS — using the technology and the research expertise of Project OIDA, the KVA-accelerated venture working on epistemic knowledge for the AI era. The seven stages below are the reference components. We calibrate per context.

Source ingestion

Documents, emails, meetings, CRM, code, structured databases, expert interviews. Whatever the organisation actually produces, not just whatever sits in SharePoint.

Semantic segmentation

Decomposed by claim, not by token count. The unit of knowledge is the unit of meaning, not whatever fits in a chunk.

Entity & actor resolution

Who said what, in what role, with what authority. The same sentence from a junior analyst and a board member is not the same sentence.

Knowledge classification

Each unit typed by its epistemic role. Factual claim, opinion, hypothesis, decision, commitment. The downstream AI gets to know the difference.

Provenance tracking

Every Knowledge Object anchored to its source and to its dependency chains. Always defendable, always traceable.

Confidence & uncertainty

Scored by formula, not by an LLM that read the sentence and felt strongly about it. Same input, same score, every time.

Lifecycle management

Decay, contradiction resolution, version control, archival. Knowledge ages. The system has to age with it.

LLMs are configuration interpreters. Not schema generators.

The model maps natural language to pre-validated epistemic schemas. It does not invent ODE parameters. It does not propose new epistemic classes on the fly.

Schema violations hard-fail. There is no fallback path. There is no “close enough”.

That is the only way to keep schemas uniform across instances and graph integrity intact over time. Everything else compounds into chaos.

OIDA — Epistemic Knowledge for the AI era.

OIDA is the tech company accelerated by KVA, dedicated to epistemic knowledge infrastructure. Project OIDA is its research arm — the people who formalise the framework KVA puts into clients.

The framework models organisational knowledge as four interlocking layers. It works at the epistemological layer — what is known, how confidently, what is decaying — distinct from ontological platforms (think Palantir) that model what entities exist and how they relate. The two layers are complementary. They are not in competition.

Knowledge Objects (KOs)

The atomic unit. Typed, sourced, scored. Carries everything the system needs to know about itself.

Knowledge Gravity Engine (KGE)

Importance and decay computed by ODE, not by LLM mood. Stable across queries, stable across months.

Hybrid Retrieval

Dense, sparse, and structural retrieval, weighted by what is actually being asked. None of the three on its own is enough.

Organisational Belief System (OBS)

What the company collectively believes, with what confidence, and how those beliefs are moving under new evidence.

Project OIDA position paper in preparation for H2 2026.

What a Knowledge Object actually looks like.

Two excerpts. The dataclass on one side, the API the agent layer uses to interrogate the OBS on the other. Schemas are calibrated per domain — these are the bones.

knowledge_object.py

from dataclasses import dataclass
from datetime import datetime
from typing import Literal

EpistemicType = Literal[
    "factual", "opinion", "hypothesis",
    "decision", "commitment", "observation",
]

@dataclass(frozen=True)
class KnowledgeObject:
    id: str
    claim: str
    type: EpistemicType

    # provenance
    source_id: str
    source_span: tuple[int, int]
    actor: Actor   # who, role, authority

    # epistemic state — deterministic
    confidence: float                 # [0, 1]
    decay_half_life_days: float
    created_at: datetime

    # graph relationships
    supports: list[str]
    contradicts: list[str]
    derived_from: list[str]

obs_query.py

# What does the org currently believe about X,
# and how has that belief evolved?
belief = obs.query(topic="pricing strategy")

belief.consensus
# weighted by confidence and decay state

belief.contradictions
# explicit conflicts among KOs, with sources

belief.evolution(window="18m")
# trajectory over time, decision points marked

belief.decisions_made_under(
    snapshot=obs.snapshot_at("2025-Q4"),
)
# which decisions stood on now-decayed beliefs?

OBS — what the organisation thinks, in writing.

The Organisational Belief System is the layer that turns AI from a productivity tool into infrastructure. KVA structures it for each client during the engagement.

It answers questions no current enterprise system can:

01
What does this organisation currently believe about X — and how confidently?
02
How has that belief shifted in the last eighteen months?
03
Which decisions were made under which beliefs, and have those beliefs since decayed?
04
Where do internal experts disagree — and what evidence are they each holding?

Where it is already running.

Production

Epistemic Ingestion Pipeline

Turns raw sources into structured knowledge candidates and keeps a hard line between what the LLM proposes and what the system is allowed to write. The boundary is the point.

Internal

KVA AI Research Hub

Our own research and operational knowledge — papers, dossiers, implementation notes — sit on the same substrate we ship to clients. Eat your own cooking.

Client

Enterprise architectures

Sales enablement, compliance memory, R&D, legal precedent, investment knowledge. Same substrate, calibrated to domain, owned by the client.

What an engagement actually produces.

Knowledge architecture audit — current epistemic state

Knowledge Object schema, calibrated to your domain

Ingestion pipeline with epistemic governance

Validation Gate — hard-fail, no fallback

OBS instantiation — your organisation, modelled

Retrieval and reasoning interfaces

Hand-off to the agentic layer

Map your knowledge before someone else writes the wrong story about it.

Most engagements start with a focused audit of the knowledge surface. From there we design the pipeline, instantiate the OBS, and hand off to the agentic layer.

Map your architecture →Next: Agentic Systems →