From Shannon to Structure.
Claude Shannon defined communication as the transmission of information across a channel with noise. Drop extends this: it measures not only transmission, but comprehension — the reduction of uncertainty in human meaning.
Structure communication, and it becomes computable.
Make it computable, and you can measure it.
Measure it, and you can improve it.
Drop reframes communication as a directed knowledge graph — making organisational thinking visible, traceable, and measurable.
Every message becomes a node, every transmission becomes an edge, and understanding becomes measurable across the entire network. This structure allows Drop to apply network science and machine learning to model communication efficiency, identify bottlenecks, and infer comprehension.
Nodes
People, teams, systems, or documents — any entity that participates in communication.
Edges
Transmission links — who shared what, with whom, and when.
Weights
Interaction strength, latency, and comprehension level.
Measuring understanding, not just attention.
Traditional communication metrics measure attention — views, clicks, opens. Drop measures understanding: how comprehension propagates, transforms, and decays across your organisation.
The speed of understanding spread — how quickly comprehension propagates through your organisation.
Coverage completeness — the percentage of your target audience that actually received and engaged.
Semantic match between intent and comprehension — are people understanding what you meant?
How quickly clarity decays after transmission — the shelf-life of understanding.
Organisations are not hierarchies of people; they are networks of communication.
Understanding these networks — how ideas move, stall, and mutate — is the key to organisational intelligence. Drop applies principles from network science to measure how comprehension propagates.
Reach
Degree Centrality
How connected a communicator is — the number of edges in and out of each node.
Influence
Betweenness Centrality
Bottleneck or bridge potential — how many shortest paths pass through someone.
Access
Closeness Centrality
How quickly someone can reach the entire organisation — average distance to all nodes.
Authority
Eigenvector Centrality
Who influences the influencers — recursive importance based on connection quality.
Cohesion
Clustering Coefficient
How tightly-knit a group is — do the people you communicate with also communicate with each other?
Dialogue
Reciprocity Rate
Is communication two-way or broadcast? The ratio of bidirectional to total edges.
Three fundamental abstractions power the Drop system.
Drops
Structured, multimedia capsules of communication. The atoms of understanding — each carrying metadata, semantic embeddings, and behavioural data.
Drop Types
- • Expert Drops — codify deep domain knowledge
- • Briefing Drops — deliver updates and strategy
- • Onboarding Drops — teach process, culture, and tools
Flows
The pathways through which Drops move. The bloodstream of information — tracking who receives, understands, and forwards each message.
Flow Directions
- • Top-Down — leadership → teams
- • Bottom-Up — teams → leadership
- • Cross-Functional — team ↔ team
Intelligence
The observation and optimisation layer. The nervous system of the organisation — measuring, learning, and improving how understanding happens.
Intelligence Captures
- • Behavioural data — views, completions, reactions
- • Flow metrics — velocity, bottlenecks, centrality
- • Comprehension — dwell time, Q&A density
The first model built to see how understanding moves—and where it stops.
A new class of semantic models that understand how messages are sent, received, and actually understood across organisations. Signal-1 builds a live semantic model of how your organisation communicates — how messages travel, mutate, drift, and are actually understood.
It models behaviour, not just language. It's not a writing model. It's an understanding model.
Ingest
Messages, docs, decks, Slack, transcripts
Semantic Encoding
Structure, claims, tasks, context
Propagation Model
Audience modelling, drift prediction
Clarity Engine
Rewriting, optimisation, routing
Understanding Graph
Persistent meaning over time
AI as orchestration, not decoration. Drop treats AI as systemic intelligence.
Drop's AI doesn't replace communicators — it amplifies clarity, ensures distribution, and closes feedback loops. By encoding communication as data, Drop allows AI to reason across the entire lifecycle: creation → delivery → comprehension → adaptation.
Input Understanding
Interprets raw text, audio, or slides from creators using LLMs, ASR, and semantic segmentation.
Knowledge Structuring
Breaks input into Drop-ready blocks with outline, summary, tone, and metadata.
Generation
Synthesises text, voice, visuals, and video using multi-modal transformer models.
Embedding & Search
Produces vector representations for retrieval, RAG, and semantic clustering.
Orchestration
Manages dependencies between models and workflows using intelligent pipelines.
Analytics & Feedback
Aggregates usage and comprehension data for continuous model optimisation.
Agentic Loops
Summariser
Converts meetings & docs into structured Drops
Q&A Agent
Answers context-aware questions inside Drops
Flow Optimiser
Reorders Drops for better coverage
Insight Synthesiser
Generates periodic intelligence reports
Comprehension Agent
Predicts understanding via behavioural signals
The mathematics of meaning.
Drop seeks to maximise:
Uv = ΔC × ΔR / Δt
Understanding Velocity — the rate at which clarity and reach compound over time.
And minimise:
H = −Σ pi log pi
Information Entropy — variance in how uniformly a message is understood across its audience.
Drop gives organisations a new layer of cognition — a system that doesn't just store or send information, but learns how understanding happens.
Experience the science of understanding.
Drop is building the infrastructure for how communication moves inside modern organisations. Join the waitlist to be first in line.