New Kerovan for tokenized funds. Free pilot from mid November. See Kerovan →

Kerovan Public Beta for Tokenized Fund Operations

Mission-Critical Industries

November 11, 2026

The Bottom Line

Kerovan is a new Scalytics product for tokenized funds, bonds and deposits. It keeps on-chain tokens and the off-chain legal register and bank ledgers in step, and proves what happened. The public beta is open.

Kerovan reconciles both books, holds anything that looks wrong until a person approves it, and writes every decision to a signed audit log. It runs on your own infrastructure, on premise or in a hybrid cloud, as a plugin to the Kafka you already run. It is read-only and needs no core banking changes. The beta is a free pilot on your own infrastructure that later turns into a paid subscription.

‍

One fund lives in two books

A tokenized fund unit exists twice. The token sits on chain. The legal register sits with the transfer agent, and the cash sits on a bank ledger. Regulators treat the off-chain register as the legal record and expect the two to match.

Someone has to keep them in step and prove it. Today teams chase the gap by hand.

‍

Where the books drift apart

The two books do not stay in step on their own. Three things go wrong.

  1. A transfer settles on chain but not on the register.
  2. A ledger entry has no token behind it.
  3. NAV and price data is self-reported, so nobody can show where the number came from.

The third one is easy to miss. Chainlink's own documentation says reserve data reported by an issuer's self-hosted API "carries additional risks". The weak spot is not the oracle. It is inside the administrator, where the number is made.

Then the auditor asks what broke, what you decided and who approved it. Under DORA, operations teams have to answer that. Emails and spreadsheets do not hold up.

‍

Why Kerovan exists

Banks already run Kafka, Redpanda or AWS Kinesis. The events are there.

What is missing is the link between the two books. The token books sit on chain. The legal register and the ledgers sit off chain. Nothing in that stack connects them, checks them against each other and proves the result.

Kerovan is that link, on the Kafka you already run. It plugs into Kafka (Confluent, Amazon MSK, Strimzi), Redpanda or AWS Kinesis. No Kafka? It reads change events from a database read replica instead.

‍

How Kerovan works

Kerovan runs in six steps, all inside your own infrastructure: on-prem, hybrid cloud, or wherever you want.

Here is one break, end to end. A token transfer is finalized on chain, but the receiving holder is not on the legal register. Kerovan holds any follow-up transfers and asks for four-eyes approval. The operations lead and the compliance lead both sign, and the decision and both approvals go into the audit log. A signed instruction then goes to custody, and the report lands on the Kerovan topic.

Kerovan does the checking. Your people make the calls that matter.

1

Sense

Ingests and standardizes events from blockchains, legal registers, and bank ledgers into a single unified format.

Data Ingestion
2

Reconcile

Continuously compares books using only finalized chain data, ensuring reorgs do not trigger false breaks.

Continuous Checking
3

Decide

Automates low-risk fixes. Any hold, freeze, or pause triggers a mandatory requirement for signed four-eyes approval.

Human in the Loop
4

Prove

Writes every single automated decision and human approval into a permanent, signed, hash-chained audit log.

DORA Compliance
5

Publish

Signs accurate NAV and price data precisely at the source before handing it off to your chosen oracle.

Verifiable Oracles
6

Act Safely

Holds no production keys. It only sends signed instructions to your existing custody system to execute the final action.

Zero Key Risk

‍

A plugin for your Kafka, not an integration project

Kerovan is a plugin, designed with partners, for Kafka (Confluent, Amazon MSK, Strimzi), Redpanda and AWS Kinesis. Your own mirroring copies an allowlist of topics, read-only, into a Lascaris cluster inside your infrastructure. On Confluent, a mirror topic is a read-only topic. You pull signed reports from a Kerovan topic. That is the whole connection.

  • No core banking changes and no new APIs.
  • Your data never leaves your infrastructure, and nothing writes back.
  • A smaller security review: read-only topics, no keys, no inbound writes.

‍

No Kafka? Kerovan can also read change events from a database read replica.

‍

The chains and oracle you already use

You do not change chain or oracle to use Kerovan.

  • One EVM connector covers Ethereum, Base, Arbitrum, Polygon, Avalanche, BNB Chain, Hedera and private Besu networks.
  • Canton is supported for institutions that issue or settle on it. Solana and Stellar on request.
  • Kerovan signs NAV and price data at the source. Your oracle carries it on chain. It complements your oracle and does not replace it.

‍

Who it is for

Kerovan is for the people who answer for the books: heads of fund operations, reconciliation, risk and compliance.

  • Fund administrators sign NAV and price data where it is made and show where every number came from.
  • Transfer agents see every gap between token holders and the legal register, with the reason and the approval.
  • Custodians and banks reconcile tokenized deposits and fund units against their ledgers without touching core banking.

We start in the EU and US. We work with banks to develop Kerovan further.

‍

The public beta

The beta is a free pilot that runs on your own infrastructure and later turns into a paid subscription.

You mirror an allowlist of topics into Lascaris on your own infrastructure. You see reconciliation runs, a break held by the circuit breaker, the four-eyes approval and the audit entry behind it. Your data, your keys and your decisions stay with you.

If you run fund operations, reconciliation or compliance for tokenized funds, bonds or deposits, see Kerovan and book a demo. For the operations view, read tokenized fund operations on the live register.

About Scalytics

Scalytics builds Lascaris, the platform that runs governed AI agents on the Kafka you already have. We help banks, insurers and fund teams put AI into production without moving customer data and without losing the record of what each agent decided.

Our founding team created Apache Wayang, the open source framework that runs computation where the data lives instead of copying it around. The same idea is at the heart of Lascaris.

Our mission is to keep data in place, bring the compute to the data and give every regulated team AI they can defend in front of an auditor.

The experts in mission-critical data and AI.

Bring us your hardest problem. We'll scope it with you.