tacit
Organisational Intelligence for the AI Era
the problem
The $1.8 Trillion Blind Spot the problem
$1.8T
Global AI investment by 2030
But 85% of AI projects never reach production.
The failure isn't technology. It's knowledge.
Sources: Gartner 2025 · BCG 2025 · McKinsey Global AI Survey 2025
The Numbers That Should Terrify Every Board the evidence
88% → 5%
Adopted AI — see real returns
BCG + McKinsey, 2025
84%
Haven't redesigned jobs for AI
Deloitte State of AI, 2025
30%
GenAI projects abandoned after POC
Gartner, 2025
10,000
Baby boomers retire daily in US + UK
Bureau of Labor Statistics
The Missing Layer the gap
Your best engineer diagnoses a failure before the logs catch up.
Your senior analyst sees the pattern three weeks before the model.
Your veteran operator knows which step to skip — and why it still works.
None of this is written down. When they leave, it leaves with them.
This is tacit knowledge.
It's your real competitive advantage.
The Knowledge Pyramid the framework
Explicit
Documents · RAG · Policies
Where everyone plays
10%
Implicit
Experience · Skills · Know-how
30%
Tacit
Pattern Recognition · Intuition · Heuristics
Where WE play
60%
Your AI agents only use the top layer.
Why Now — The Convergence the urgency
Retirement Wave
50% of claims adjusters, 25% of underwriters retire within 10
years. Training a replacement takes 5–10 years.
AI Budget Explosion
$1.8T by 2030 — spent on engines without fuel. 82% of
leaders prioritise AI, only 22% deployed at scale.
Foundation Models
LLMs can now parse natural language heuristics and classify
decision patterns. The technology bottleneck has lifted.
Regulatory Pressure
EU AI Act, Solvency II, IFRS 17 mandate explainable AI decisions.
Undocumented expert logic = compliance gap.
the solution
Tacit is Organisational Cognition Infrastructure.
We capture, codify, and operationalise the tacit knowledge that makes your organisation work — the
invisible intelligence that no system has ever reached.
Before you can manage knowledge, you have to capture it.
Before RAG can retrieve expert judgment, someone has to extract it.
We are the layer before the layer.
Three Capture Paths — One Knowledge Graph the platform
Path A
Passive Behavioral Capture
Observer SDK watches how experts work. Captures overrides, hesitations,
attention patterns.
Path B
Structured Interview Agent (SIA)
AI-powered conversational agent conducts deep dialogues. Extracts
reasoning and exceptions.
Path C
Contrastive Expertise Analysis
Compares expert vs novice cohorts statistically. Cohen's d effect sizes
identify differential behavior.
THE KNOWLEDGE
GRAPH (Neo4j)
The 100-Marker Cognitive Framework the science
A systematic taxonomy of expert cognition across 7 layers — no one else has
built this.
100 markers total · 8 capture
strategies implemented
It's Not Just Theory. It's Built. traction
16
API Routers
65+
REST Endpoints
20
Neo4j Node Types
34
Graph Constraints
14
Service Modules
19
Dashboard Pages
8
Capture Strategies
3
Extraction Paths
Pure Python · Zero vendor lock-in · Deployable today
The Knowledge Graph — Live the engine
Every captured expert session becomes a rich subgraph of connected intelligence.
Expert
→
Session
→
Event
→
Pattern
→
Heuristic
KnowledgeNugget
Outcome
DriftAlert
GapAnalysis
Archetype
Confidence scores — every heuristic has a weighted
confidence (0–1) based on evidence
Decay rates — knowledge not reinforced automatically
fades over time
Freshness timestamps — when was each heuristic last
observed in the wild?
Knowledge that isn't reinforced fades. Knowledge that's validated strengthens.
Zero Friction — Deploy in Days deployment
01
NPM Package
Install via npm for modern web apps
02
Script Injection
Single <script> tag added to any page
03
Tag Manager
GTM / Tealium zero-code deployment
04
Browser Extension
Enterprise-managed Chrome extension
0
API integrations required
<2KB
Bandwidth per minute
100%
PII redacted at browser layer
the products
Not One Product — A Platform product suite
Tacit Preserve
Expert Knowledge Preservation
Captures retiring experts' decision logic before they leave.
The knowledge insurance policy
Tacit Transfer
Accelerated Knowledge Transfer
Converts senior expertise into structured onboarding for juniors.
Cut ramp-up from 5 years to 3
Tacit Align
AI Model Alignment
Feeds captured heuristics into AI models via Intent Engine API.
Model accuracy 74% → 86%
Tacit Decode
Behavioral Intelligence
Apply capture methodology to model customer decision patterns.
Demographic → Cognitive segmentation
One Capture — Five Value Streams value creation
Expert
Knowledge
Graph
Knowledge
Graph
AI Models
Better predictions
Training Programs
Faster onboarding
Compliance Systems
Auditable decisions
Customer Intelligence
Behavioral segmentation
Cross-Org Benchmarking
Anonymised industry patterns
That's platform economics.
What We Are / What We Are Not positioning
| Category | What They Do | What We Do |
|---|---|---|
| Knowledge Management | Confluence captures what people write down. | We capture what they can't articulate. |
| RAG | Retrieves existing documents. | We create what RAG should be retrieving. |
| Process Mining | Celonis tells you WHAT. | We tell you WHY. |
| Decision Intelligence | Palantir starts top-down with data. | We start bottom-up with people. |
The first compiler for human decision-making.
beachhead market
Insurance Underwriting beachhead market
The purest tacit knowledge function in any industry.
9.15 / 10
Vertical Score (highest of 6 evaluated)
$410M → $7.9B
Market 2025 → 2033
44.7%
CAGR
82%
Prioritise AI
“
Your best underwriter looks at a submission and sees things nobody else sees. She knows that restaurants
with valet parking in flood zones are actually lower risk — because the valet moves cars before
storms.
She knows that a manufacturer with ISO 9001 but no ISO 14001 is a red flag — it signals
cost-cutting on environmental controls, which correlates with safety incidents.
She knows 300 things like this.
None of it is in your underwriting guidelines.
And she retires in 18 months.
None of it is in your underwriting guidelines.
And she retires in 18 months.
Three Doors Into Every Carrier the buyer
Primary
Chief Underwriting Officer
“I'm losing 30% of my senior
underwriters in 5 years. Juniors can't replicate their judgment.”
Secondary
CTO / CDO
“We've deployed AI for extraction and
quoting. Risk assessment quality still depends on individuals.”
Sponsor
Chief Risk Officer
“Regulators want us to explain how we
make risk decisions. Undocumented expert logic = governance gap.”
What We've Proven live system data
Sarah Chen · 15yr
Senior (Tier 3)
Marcus Rivera · 7yr
Mid (Tier 2)
Priya Patel · 2yr
Junior (Tier 1)
54
Sessions captured
722
Behavioral events
8
Heuristics extracted
9
Patterns detected
4
Gap analyses (Cohen's d: 0.92–1.82)
16
Of 100 markers hit
Built and running. This is a working system.
the business
Scaling Sequence go-to-market
Phase 1 · 0-12mo
Insurance Underwriting
3–5 mid-market carriers. Build the playbook.
Phase 2 · 6-18mo
Financial Services
AML/KYC — same buyer profile, same capture mechanics.
Phase 3 · 12-24mo
Manufacturing
$3.3B → $39.2B market. Expert-heavy, process-dense.
Phase 4 · 18-30mo
Pharma Regulatory
Highest deal sizes. Deep domain expertise required.
Land and Expand business model
Observe
£60K/yr
+ £1,200/expert
Behavioral capture
Passive observation
Basic analytics
Passive observation
Basic analytics
Engage
£150K/yr
+ £3,000/expert
Behavioral capture
+ SIA interviews
+ Knowledge graph
+ SIA interviews
+ Knowledge graph
Transform
£350K/yr
+ £5,000/expert
Full platform
+ Intent engine
+ Compliance + Audit
+ Intent engine
+ Compliance + Audit
Mid-market: $200K–$400K | Enterprise: $500K–$2M
Expansion within existing clients: 2–3x initial deal
The Flywheel defensibility
Pattern
Library
Library
Capture expertise
at Client A
at Client A
Build methodology
for function
for function
Deploy faster
at Client B
at Client B
Expand to next
function
function
Each client makes the platform smarter for the next
After 20 clients: the most comprehensive library of how enterprises make decisions.
Four Defensible IP Pillars intellectual property
Algorithmic IP
The Heuristic Compiler
First compiler for human decision-making. Multi-modal ML extracts
heuristics from raw behavioral telemetry.
Architectural IP
Shadow Context Stitching
Zero-integration cross-app capture. OS-level observation stitches
context across siloed enterprise apps.
Data Structure IP
Experiential Decay Graphs
Self-pruning temporal knowledge graphs. Nodes possess experiential
half-lives based on behavioral reinforcement.
Generative IP
Synthetic Expert Simulation
Cognitive clones of top performers. Generate persona-specific,
explainable decision rationales.
Provisional IP architecture marked as trade secrets.
Defensive patenting post-Seed.
Market Opportunity market
We sit at the intersection of three explosive markets.
Sources: Grand View Research 2025, MarketsandMarkets, Allied Market Research
the ask
Pre-Seed the ask
£750K
50%
Engineering
Platform hardening, ML pipeline
25%
Insurance Pilots
First 3 carrier deployments
15%
Go-to-Market
Insurance vertical sales
10%
Operations
Legal, compliance, ops
Capacity-limited by design.
We partner deeply with 3–5 enterprises at a time.
No surface-level deployments.
The billion-dollar company is not ‘tacit knowledge
capture.’
It is not ‘AI calibration.’
It is the infrastructure that makes the invisible intelligence of organisations visible, durable, and computable.
That has never existed before.
We build it.
It is not ‘AI calibration.’
It is the infrastructure that makes the invisible intelligence of organisations visible, durable, and computable.
That has never existed before.
We build it.
hello@tacitlabs.ai · tacitlabs.ai
© 2026 Tacit Labs. Organisational Cognition Infrastructure.