Now open Pilot program for neobanks, fintechs and exchanges

The immune system
for digital finance.

Agentic AI that detects, decides, and defends.

ISOFORT runs autonomous AI agents that watch every transaction, on-chain and off, from onboarding to settlement. They investigate the risk, weigh the signals, and act, with your team in control. We don't write rules. We build immunity.

Schedule a demo How it works ↳ Closed beta, ~5 partners per quarter
Live stream
Transaction monitoring
connected
Risk events · Last 60 minutes
Receiving live · 12,847 events / min
All 1284
Blocked 42
Review 11
Filter
Detected · 24h
1,284 ↑ 4.2%
Neutralized
1,251 ↑ 3.8%
Under review
33 ↓ 12%
False positive rate
2.6% ↓ 67%
Status Event Customer Amount Score
blocked evt_9f4a2b71
wallet drainer · on-chain
Ξ0x9c4f…a71b 12.4 ETH 0.96
blocked evt_9f4a2b70
ATO attempt · velocity
JKJ. Klimkowski € 12,400.00 0.94
step-up evt_9f4a2b6f
address poisoning · look-alike
Ξ0x441c…9c7e 2.0 ETH 0.71
blocked evt_9f4a2b6e
synthetic identity
RFR. Fernández € 5,700.00 0.88
released evt_9f4a2b6d
clean · SEPA transfer
SRS. Rosenberg € 245.00 0.08
released evt_9f4a2b6c
clean · known wallet
bc1q…8f7a 0.15 BTC 0.05
Agent: investigate evt_9f4a2b71, trace the funds across chains, and draft the SAR
Agentic AI
Autonomous agents that investigate and act
Multi-source intelligence
Every signal that matters, on-chain and off
Explainable by design
Every decision, every weight, no black box
Research-backed
Detection grounded in published methods
In private beta with European neobanks, fintechs and exchanges
Web2 + Web3

Fraud crosses both worlds. So do we.

The bank transfer and the blockchain transfer, in the same case, with the same engine. Most tools pick a side. We watch both, and read them together.

Web2 · off-chain

Traditional finance

  • Cards
  • SEPA
  • SWIFT
  • KYC
  • Sanctions
  • Device
  • Credit
IF
one case
one score
Web3 · on-chain

Ten blockchains

  • Wallets
  • DeFi
  • Taint
  • Drainers
  • Mixers
  • Tokens
  • Bridges
Same address, same customer, same case. Both worlds, one view.
How it works

Four stages, modeled on biological immunity.

Every transaction triggers the same four-stage response a biological immune system runs against pathogens. Continuously, at scale, with full explanations for your audit team.

01 · Detect

Recognize the antigen

Behavioral, device, and graph signals scored against learned baselines. No static rules to tune.

02 · Isolate

Quarantine the event

Isolation Forest separates the suspicious from the routine without supervised labels.

03 · Neutralize

Decide and act

Block, step-up, or release. Every decision returns with a human-readable explanation.

04 · Remember

Build immunity

Every confirmed event re-trains the model. Tomorrow's attack is partially neutralized today.

See the full breakdown
Agentic AI · the core of IsoFort

Not another dashboard.
An AI analyst that works the case.

Most fraud tools stop at a score and an alert. ISOFORT keeps going. Autonomous agents pick up the alert, investigate it end to end across every chain, trace the money to its source, and draft the regulatory filing. Your team approves. The agents do the work.

Investigation agent · case_4f2a running
Pulled 1,284 transfers across 4 chains
Traced taint to a sanctioned mixer
Flagged 3 look-alike counterparties
Drafted SAR narrative · 6 citations
Awaiting analyst approval · freeze request
Triage agent
Reads every alert, ranks by real risk, and clears the noise before it reaches your queue.
Investigation agent
Works a case end to end: pulls the transfers, traces the flow, writes the filing.
Monitoring agent
Watches flagged wallets around the clock and re-opens a case the moment they move.
3
autonomous agent types working in parallel
10+
chains investigated in a single pass
100%
regulated actions gated by human approval
24/7
continuous autonomous monitoring
The architecture

End-to-end immunity, in three layers.

Not another data feed or another orchestrator. The layer that decides. Intelligence flows up from every chain and data source, into an explainable scoring engine that knows which signals lead and which stay quiet, into autonomous agents that work the case, with your team in control at the top.

01
Foundation
The intelligence layer

Every major chain, read directly and normalized into one feed: balances, transfers, DeFi positions, plus the sanctions, threat, and identity signals that give each address context. Vendor-neutral, so no single source is a blind spot.

10+ blockchains RPC DeFi positions Balances & transfers Sanctions & watchlists Threat intelligence Token security Off-chain identity Device fingerprint
02
Engine
Explainable risk scoring

Hard-veto, then capped per-source layers, then an ML modifier, then Identity Hub context, with taint inherited across chains. Every point of the score traces back to the source that produced it. No black box.

Hard-veto Per-source caps ML ensemble Identity Hub Cross-chain taint Full audit trail
03
Agentic AI
Agents that work the case

Autonomous agents triage alerts, investigate end to end across on-chain and off-chain signals, and draft the regulatory filing with citations. Every file / freeze / block passes through a human-in-the-loop gate.

Triage agent Investigation agent Monitoring agent SAR drafting Human-in-the-loop MCP tools
Signal flows up · explanations flow back down
See how we decide
The platform

Modular building blocks for every transaction.

One platform, eight modules, one contract. Plug in only what you need, scale into the rest as you grow.

Identity checkflagged
Documentpassport
Livenesslow
Synth score0.83
02 · Onboarding

Synthetic ID defense

Catch fabricated, stolen, and AI-generated identities at signup.

SESSION INTEGRITY
03 · Access

Account takeover

Block credential stuffing, session hijacking, and SIM swap attempts.

04 · Cyber

Bot & agentic defense

Tell humans, bots, and autonomous agents apart at the edge.

05 · Network

Fraud ring detection

Surface hidden links between accounts, devices, and counterparties.

taintedsanctioned mixer18.4
hopbridge · cross-chain16.1
cash-outVASP · exchange15.8
06 · On-chain

On-chain forensics

Score any wallet, then trace tainted funds upstream and downstream across chains.

Learn more
Verify recipientblock
Sending to0x441c…9c7e
Looks like0x441c…3fdf
Paid beforenever
07 · Prevention

Pre-send protection

Catch address-poisoning and look-alike drainers before the payment is signed.

Learn more
donetraced taint · 4 chains
donedrafted SAR · 6 cites
waithuman approval
08 · Agentic AI

Agentic AI analysts

Autonomous agents investigate cases and draft filings, with a human-in-the-loop gate.

Learn more
Integrations

Plugs into the stack you already run.

SDKs for every major platform, REST and gRPC APIs, real-time webhooks, native connectors for payment rails and chains.

Stripe
Adyen
Bitcoin
EVM chains
Solana
SEPA
SWIFT
AWS
GCP
Azure
REST API
Webhooks
Built for

Built for every shape of finance.

From neobanks and exchanges to incumbent banks and credit unions. One immune layer that adapts to each threat model, at any scale.

Pilot program

Become an early partner.

A limited cohort of neobanks, fintechs and exchanges deploying immunity before the rest of the market knows what to ask for. Founding-customer pricing, a direct line to our engineering team, and a co-built first integration.

Schedule a demo See the pilot program
Field notes

Research from the front lines.

Research and engineering notes from the team, publishing soon. What we've learned about modern fraud, and how we think immune systems should be built.

Audit-ready by default

Built for European supervisory frameworks. Every decision is logged, explainable, and exportable.

SOC 2 in progress
ISO 27001 aligned
GDPR
PSD2 / SCA
Get started

Your platform is already
under attack.

Apply for the pilot cohort. We respond within 48 hours. No slides, no salespeople until you decide we're a fit.