BCSentinel is a future formation intelligence platform that detects formation processes before their public visibility, delivering actionable alpha to institutional investors.
Not a dashboard. Not analytics. An intelligence platform.
Method
Formation happens in three moments. We're built to detect the first one.
Markets price what's visible. By the time a launch, a listing, or a disclosure makes a formation public, it has usually been assembling on-chain for weeks or months. That's the window we detect in — before the market has priced it.
T1 · FIRST SIGNAL
A pattern starts
A multisig Safe gets deployed, a role gets granted, funding starts concentrating from a small number of sources. Individually, any one of these is noise.
T2 · ALERT THRESHOLD
The pattern corroborates
Enough independent signals line up — deployment timing, funding concentration, entity age — that the case crosses our internal threshold for human review.
T3 · PUBLIC VISIBILITY
The entity confirms itself
A launch, a listing, a press mention, a verified contract. This is where most tools start looking. We measure how far ahead of this point we actually were.
Every classification we produce has to trace back to a specific, observable piece of on-chain evidence. If we can't point to the transaction, the contract, or the timing that justifies a score, we don't assign one.
Our engines are built to say outright when they don't have a historical basis for estimating a probability — rather than estimate one anyway.
Architecture
Every decision is written down before it's built. Every engine is named, not just counted.
BCSentinel isn't a model we trained and shipped — it's a documented architecture: decision records, design specifications, and a chain of investigation engines that each closed a specific, named blind spot as we found it.
investigate_wallet — core classification engine
Now in its seventh revision. Version 3 closed the blind spot that mattered most: a sophisticated actor operating through a Gnosis Safe showed up only as execTransaction(...) — the real action hidden inside.
- Safe-payload decoding — recovers the real selector inside every Safe transaction
- MultiSend decoding — unpacks every batched sub-transaction independently
- Economic-value extraction — converts raw transfer data into readable amounts
- Contract-creation check — catches deployments with no destination address, including EIP-1167 minimal-proxy clones
Every wallet in our calibration set carries a direct, timestamped investigate_wallet output file — the pipeline's own evidence trail, generated as it runs, not a label applied after the fact.
Scope today: Ethereum mainnet (Layer 1) — 187 formations investigated in depth, with live detection already running against our calibrated signal set.
Confidence Engine — evidence reliability
A separate engine from the classifier above: this one scores how much to trust the evidence itself, not the size of the formation. Five independent categories, non-linear gating between them, and an explicit, severity-weighted penalty for contradictions — with full provenance kept on every input that fed the score.
Fraud Risk Engine — outcome classification
Classifies what actually happened to a formation after the fact: active, legitimate wind-down, confirmed exit scam, slow insider extraction, or silent shutdown. Its silence-pattern detector was validated against a real historical case before being trusted on the rest of the dataset.
Organization-first prefilter — TS-002 / TS-002B
Runs ahead of individual-wallet investigation, scanning for organizational signatures across a wider net before a candidate ever reaches the classifier. Every prefilter carries an explicit recency window, so a candidate that goes quiet doesn't sit in the queue as a false positive indefinitely.
A library only moves from draft to production once it clears a fixed validation threshold against real cases — we don't promote a pattern library on vibes. In parallel, a set of active prospective investigations is validating this architecture against formations happening right now, not just against history.
Adaptive intelligence
A platform that continuously evolves.
BCSentinel is designed to evolve continuously. Every validated formation expands the platform's institutional knowledge.
As new organizational structures appear across the on-chain ecosystem, the platform extends its understanding of Formation DNA and searches for previously unseen archetypes.
Rather than relying on a fixed analytical model, BCSentinel progressively strengthens how it recognizes, classifies and interprets emerging structures. Every confirmed case is retained as evidence — the platform learns cumulatively from what it has already resolved.
The financial industry reacts to events.BCSentinel detects formation as it emerges.
Markets price visibility.BCSentinel detects what the market cannot yet see......
Verification
We don't ask you to take our word for it.
Two independent mechanisms back every case we track — one checks the claim against public sources, the other proves we didn't rewrite our own history.
Cross-source confirmation
Every candidate we track gets checked against GitHub code search, Etherscan's verified-contract registry, and DefiLlama's protocol index — three independent, public sources outside our control. A case only advances when the on-chain pattern lines up with something checkable off-chain.
Cryptographic sealing
Starting this cycle, every new formation we begin tracking is sealed the moment we see it — a hash of the case record, chained to the one sealed before it, with a timestamp. If that formation becomes public later, anyone can recompute the hash and confirm it matches what we sealed.
Institutional markets
Built for the organizations that move institutional capital.
Investment Institutions
- Hedge Funds & Multi-Strategy Funds
- Asset Managers
- Sovereign Wealth Funds
- Pension Funds & Endowments
- Private Equity & Venture Capital
- Investment & Merchant Banks
- Specialized Family Offices
- Quantitative & Digital Asset Investment Firms
- Insurance Companies
Digital Asset Ecosystem
- Prime Brokers & OTC Trading Firms
- Market Makers
- Blockchain & Layer 1 / Layer 2 Foundations
- Protocol Foundations
- Stablecoin & Token Issuers
- Digital Asset Infrastructure Providers
Enterprise
- Tokenization Companies & RWA Originators
- Institutional Custodians
- Financial Infrastructure & Digital Asset Service Providers
- Payment Networks
- Enterprise Blockchain & Web3 Infrastructure
- Crypto Treasury Companies
- FinTech Companies
Public Sector
- Central Banks
- Ministries of Economy
- Financial & Securities Regulators
- Financial Intelligence Units
- National Treasury Departments
- Government Digital Transformation Agencies
Where we are
Built in public order, documented at every stage.
BCSentinel is developed by EMERGE as a layered build — each stage's governing documents and architecture decisions are written down before the next stage starts.
The historical validation above tells you how the methodology performs against formations whose outcome we already know. The next milestone is proving it holds in real time, in public, as formations happen — we're building that track record now, not claiming it yet.
Contact
Institutional inquiries
We review inquiries directly and respond to organizations we can usefully work with.