Real-time risk decisions, with the proof built in.
Nyx, our decision platform, tells the systems you already run whether a payment or loan should go ahead, for people and AI agents alike, and keeps the evidence.
Built for Banks · NBFCs · Fintechs · Payment Processors · Lenders · Merchants
- A person
- An AI agent
- A merchantOn a mandate
- IssuersPaying side
- PSPs and acquirersGetting paid
- LendersPARTNER-LED
- 1Verify, in build
- 2Know who, in build
- 3Decide, pilot ready
- 4Answer, pilot ready
What a dispute or an examiner reads.
Deciding is not the same as Defending. Every decision carries the evidence to defend it.
A decision on its own is no longer enough. Weeks later someone asks who acted, on whose behalf, and why it was allowed. Here is one payment, from the second it arrives to the day it is questioned.
-
01 · Day 0 · 10:42:18The input
A payment arrives.
A UPI transfer to a payee this customer has never paid.
UPIINR 38,000.00new payeeCardINR 2,340.00ApproveUPIINR 450.00Approve -
02 · Same secondPILOT READY
Nyx decides, in the authorization path.
Review, and warn the customer before the payment completes.
ReviewWarn before payingRule firedupi_new_payee_velocity -
03 · Same secondIN BUILD
The record is written beside it.
Who acted, for whom, and what they were shown.
ShownScam warningConsentConfirmed after itOutcomePaid -
04 · Day 12PARTNER-LED
Someone asks why.
“I was tricked into paying.” The answer comes from the record.
- The decision and its reasons
- The warning they were shown
- Their confirmation, with its time
Answered fromdec_EXAMPLE…
† Design target to validate in the first pilot, not yet independently benchmarked.
One engine, different decisions. Each with an honest status.
From fraud to agent verification to credit, the same engine decides and keeps the evidence for every decision.
Decide before the money moves.
Approve, review or decline a payment or an account action, in the authorization path.
Run a pilot on your data →Know who is acting, for whom, and within what limits.
Verify the agent’s signature, bind it to its principal, and check the mandate’s scope, caps and expiry.
Agentic Commerce →Every other decision, on the same engine.
Each pack is built with the first partner who needs it, and each carries its own status.
All risk decisions →Beside your systems, not in place of them.
Your systems keep authorizing and moving the money. They ask Nyx one question and act on the answer. Explore the platform →
- Model
- Managed by us
- Who runs it
- NaxVerse, single-tenant per customer
- Where data sits
- In the cloud region you choose
- Updates
- Monitoring, updates and backups are ours
- Model
- Your cloud
- Who runs it
- You, inside your own VPC
- Where data sits
- In your account, under your key management
- Updates
- We ship the release and support it
- Model
- Your data center
- Who runs it
- You, on Kubernetes on your hardware
- Where data sits
- Never leaves; a decision needs no egress
- Updates
- We ship the release and support it
The questions that come up first.
The detail behind the page, with the same qualifiers as the rest of the site.
What is Nyx?
Nyx is the product, and NaxVerse (NaxVerse Technologies Pvt Ltd) is the company that builds it. Nyx is the trust layer for decisions that move money. It approves, reviews or declines a transaction or an account action in real time, inside the authorization path, whether a person or an AI agent is acting, and attaches the evidence to defend that decision afterwards. One engine covers fraud, agent trust, credit risk, underwriting, disputes, AML and KYC through versioned domain packs, instead of one vendor per decision.
Is Nyx only for fraud?
No. Fraud management is the first domain pack and the one ready to pilot today. The engine underneath is domain-agnostic, and versioned packs specialize it for other decisions, each with its own status: agent trust is in build, and credit risk, underwriting, disputes and AML are partner-led. Every pack writes the same canonical evidence record, which is in build.
What is the evidence that gets attached to a decision?
Every decision returns the action, the score, the severity band, the policies and rules that fired, the model version and an audit reference in the same response. A canonical evidence record is written beside it: actor, principal, mandate, what was shown, consent, decision, outcome and dispute status. That record is what a dispute or an examination is answered with later. The record is in build, and its schema is versioned from the first write.
Who is it for?
Banks, NBFCs, digital banks, fintechs, payment aggregators and PSPs, acquirers, lenders and merchants. Institutions are the direct buyer. Merchants are reached through the PSPs that embed the trust endpoint for them, and a merchant can also come directly. Size is not a gate, because the hosted deployment removes the need for an in-house platform team.
What does Nyx do when the buyer is an AI agent?
It verifies the signature the payment rail presents (RFC 9421 via Visa Trusted Agent Protocol, Mastercard Agent Pay and the Universal Commerce Protocol; NPCI's Unified Agent Protocol once it clears RBI approval), resolves the principal that delegated the authority, checks the transaction against the mandate's scope, caps and expiry, baselines agent-operator behavior separately from human behavior, and returns an explained verdict with its audit reference. This is in build.
Does Nyx replace a PSP, a commerce platform or a fraud vendor?
No. The catalog of record stays where it is and money moves on the existing rails. Nyx does not replace an authorization system, a switch or a loan origination system either: those systems ask, Nyx answers with the decision and its reasons, and they act on it. Device, behavioral and consortium signal vendors are inputs to the decision rather than competitors. Nyx makes the decision and keeps the evidence for it.
Is Nyx running in production today?
No. NaxVerse is pre-launch and founder-led. The fraud pack is built and ready to pilot, agent trust is in build, and the performance figures on this site are design targets to validate in a first pilot rather than benchmarked results. NaxVerse publishes targets, not testimonials it does not have.
Where does the data live?
In your jurisdiction, in all three deployment models: hosted and managed by NaxVerse in the region you choose, inside your own cloud account and VPC, or in your data center. The engine, the API, the packs and the evidence record are identical in all three. Models train on your data only, and no deployment requires bulk extraction.
Which regulations does it address?
In India, the RBI Authentication Directions 2025 (risk-based authentication for cross-border card-not-present from 1 October 2026), the Digital Payments Intelligence Platform live since 1 September 2026 with Phase 2 scoring transactions before completion, the RBI Outsourcing Directions of November 2025 and DPDP obligations from 13 May 2027. In the UAE, CBUAE Notice 3057 and CBUAE's AI guidance of 11 February 2026. Mapped alongside: EU Verification of Payee and PSD3/PSR from 2028, UK APP reimbursement since 7 October 2024, MAS in Singapore, Australia's Scams Prevention Act 2025 and SAMA's Counter-Fraud Framework in force since 13 April 2026. Nyx deploys in any jurisdiction, and each regime is mapped with the first partner there.
How is Nyx priced?
Nyx is paid for decisions and for evidence: per scored event, volume tiered, or an annual licence for an on-premises deployment. It never takes a share of a transaction. Figures are not published; they are discussed in a conversation.
How does an engagement start?
With a shadow replay of your own decision history, run in the deployment model you choose with nothing leaving your perimeter. It produces four numbers you do not have today: false declines recoverable, mule accounts missed, agent-originated traffic share, and dispute cases defensible with evidence.
Replay your history. See what Nyx would have decided.
A shadow replay of your own decision history, in the deployment model you choose, with nothing leaving your perimeter.
Not a bank, or not the fraud use case? Bring us the decision you need to make.
Founder-led by Umesh Kolhe: two decades across the payments lifecycle at JPMorgan, TSYS and Pine Labs, with a senior team that has built and operated these systems inside banks and payment processors.
NaxVerse Technologies Pvt Ltd · Noida, India · pre-launch · The company → · Press & due-diligence pack →
HELLO@NAXVERSE.AI · FOUNDER-LED, PROCUREMENT-READY