We find the structure in complex networks.
E2N Networks is a boutique research consultancy. We analyze social, financial, and blockchain networks with graph data science — and run the applied AI systems that make the analysis possible, on hardware we own.
Capabilities that end in a decision
Four things we do well, applied to problems where the relationships between things are the point.
Network Analysis & Graph Data Science
Centrality, pathfinding, and community detection on social, financial, IT, and logistics networks. We surface the brokers, clusters, and bottlenecks that tables can’t show you.
Blockchain & Transaction Forensics
Wallet clustering, attribution, and flow analysis. We trace how funds move through a transaction graph — and what those patterns mean for an investigation or a research question.
Applied AI & Machine Learning
Embeddings, classification, and LLM pipelines applied to graph-structured data — from entity resolution to recommendation. Models are a tool, not the answer.
Self-Hosted AI Infrastructure
A private inference lab: open-weight models served on our own hardware, inside our network. Sensitive client data never leaves the building — and heavy analysis runs without a per-token meter.
Where it applies
A research process, not a black box
Every engagement follows the same three steps — scoped tightly, run reproducibly, and delivered in plain language.
Frame the question
We start with the decision you’re trying to make, not the dataset. Scope, data access, and confidentiality are settled up front.
Run the research
Analysis runs on our systems with a reproducible pipeline. You get interim findings as they develop — not a black box at the end.
Deliver and hand off
Findings, visuals, and the code. Hand-off can include the pipeline itself: we don’t lock you into us.
Sensitive data stays in the room
Confidential work runs on our on-premises inference lab — a private stack with no third-party APIs in the loop. If the data can’t leave the building, the analysis can still happen.
What we work with
Analysis
Neo4j · Graph Data Science · centrality & pathfinding · community detection · embeddings · clustering · recommender systems
Engineering
Python · SQL / PostgreSQL · pgvector · reproducible pipelines · data visualization · web applications
Infrastructure
vLLM · llama.cpp · quantization (EXL3 / GGUF) · GPU benchmarking & power tuning · on-prem serving
From the blog
Write-ups on network analysis, applied AI, and running models on your own hardware — what we’re working on and what we’re learning.
Tell us what you’re trying to understand.
A short description of the problem is enough. We’ll follow up with questions, and if it’s a fit, a scoped proposal.