Nominal
Nominal is an agentic AI platform that runs accounting work on top of an existing ERP: bank and account reconciliation, transaction matching, intercompany, consolidation and variance analysis, with human approval at every step. Pricing is not published.
What is Nominal?
Nominal is an agentic AI platform for corporate accounting. It sits on top of the systems a finance team already runs, an ERP with its subledgers, banks and procurement systems, and executes the work that normally falls outside them: bank and account reconciliation, transaction matching across entities, intercompany operations, multi-entity consolidation, variance analysis and journal entries. The company calls this category Agentic Performance Management, a term it coined, and positions it against three neighbours: the ERP as system of record, EPM tools for planning, and close-management tools that track tasks without doing them.
The architecture published on the platform page has four layers. Data connects the existing systems through an API, an on-premise deployment or data connectors. A shadow ledger mirrors the general ledger structure and unifies entities, systems and currencies in one bi-directional record. Always-on agents run the work continuously, described as deterministic where matching and reconciliation are concerned, which matters in an audited process. A close-management workspace is where agents and people meet, with approval workflows, a full audit trail and feedback loops.
Five families of agents are documented individually. Bank Reconciliation Agents ingest bank feeds and match them to the ledger even when descriptions, timing or formats do not line up. Transaction Matching Agents reconcile across ledgers, entities and systems. Transaction Patrol Agents watch for anomalies, duplicate payments, unexpected fees and unauthorised activity. Flux Analysis Agents explain what changed in a balance and why. Trigger Agents launch downstream workflows when a threshold is crossed.
The company was founded by Guy Leibovitz and Golan Kopichinsky, who previously built Cognigo, a data security company acquired by NetApp in 2019. It has raised USD 30 million, including a USD 20 million Series A led by Next47 with Workday Ventures. Customers named on the site include Leanpay, Kunai and Green Street Power Partners, and press releases add Jiffy Lube and GoGlobal Travel. What the site does not show matters too: no pricing, no self-service access, and contractual documents that cover the marketing website rather than the product itself.
What it does
- Reconcile bank activity against the general ledger continuously rather than at month-end
- Match transactions across ledgers, entities and systems, intercompany included
- Catch anomalies, duplicate payments, unexpected fees and unauthorised activity before close
- Explain balance variances and produce flux narratives
- Consolidate multiple entities, books and currencies
- Recommend and generate the journal entries missing from a reconciliation
- Trigger downstream workflows when a threshold is crossed or data shifts
When to use Nominal / When not to
A quick filter to help you decide if Nominal is the right fit.
When to use Nominal
- Controllers and accounting managers at multi-entity companies whose close depends on manual reconciliation
- Finance teams that want to keep their existing ERP (Workday, Oracle, SAP, Sage Intacct, Dynamics 365, NetSuite) rather than replace it
- Companies consolidating several legal entities, ledgers and currencies every month
- Organisations preparing for scale or an IPO that need a repeatable, auditable close
- Accounting teams absorbing more volume without adding headcount
When not to use Nominal
- Small businesses and independents: every reference on the site is mid-market or enterprise
- Buyers who need to see a price before a sales conversation, since none is published anywhere
- Anyone wanting to test the product alone: there is no free plan, no trial and no self-service signup
- Teams without a structured ERP or general ledger for the agents to work against
- Organisations with strict EU data requirements, as the site publishes no GDPR framework, no DPA, no hosting location and no subprocessor list
How to use Nominal
A typical end-to-end flow, from setup to results.
- Request a demo through the form, which is the only entry point to the product
- Go through the discovery conversation on your ERP, entities and close process
- Connect the source systems through an API, an on-premise deployment or data connectors
- Let the transactions, balances and journal entries replicate into the shadow ledger
- Have the relevant agents set up for your reconciliation, matching, intercompany or consolidation workflows
- Let the agents run continuously instead of waiting for period end
- Work the exceptions the agents surface in the close-management workspace
- Approve each step, with every action logged in the audit trail
- Feed corrections back so matching accuracy improves cycle after cycle
- Log in day to day at app.nominal.so
Pros & Cons
Pros
- Runs alongside the existing ERP, with no migration or re-platforming required
- Functional scope documented page by page: five agent families, four use cases, six ERP paths
- Human control is explicit, with approval workflows, a full audit trail and a read-only mode
- SOC 1 Type 2 announced in a dated press release, which is rarer than SOC 2 alone and relevant to public accounting
- Founding team with a prior exit, Cognigo having been acquired by NetApp in 2019
- Named institutional investors, including Workday Ventures, publisher of one of the integrated ERPs
- Multi-entity and multi-currency handled natively, with customer testimonials given by name, role and company
Cons
- No pricing at all: no plan, no tier, no order of magnitude anywhere on the site
- No trial and no self-service access, so evaluation requires a sales process
- The privacy policy covers the marketing website only, and stays silent on the data the product processes
- No GDPR mention, no data processing agreement, no subprocessor list and no hosting country published
- Nothing published on whether customer data may train models, or on any way to exclude it
- SOC 1 and SOC 2 claimed on a marketing page, with no report or trust portal available online
- Impact figures given without methodology, and no public technical documentation or API to assess the product independently
Pricing & Plans
No pricing is published. Nominal displays neither a free plan, nor a free trial, nor any price point: there is no pricing page in the navigation or the footer, and a search across every page collected returned no amount, tier or currency. The only commercial route is a demo, after which terms are negotiated. Buyers should expect an enterprise quotation rather than a published rate.
Data, GDPR & hosting
A consolidated view of how Nominal handles your data.
GDPR overview
There is no mention of the GDPR anywhere on the site. The acronym appears in none of the pages collected, including the privacy policy and the terms of use, both dated 27 March 2026. It is this silence, rather than any explicit refusal, that leads this record to mark GDPR compliance as no. What does exist is narrower: the policy offers access to the personal data held, correction or deletion on request, and an opt-out from marketing, all exercised by writing to privacy@nominal.so. There is no named data protection officer, no Article 27 EU representative, no data processing agreement, no subprocessor list, no stated legal basis and no transfer mechanism, and the terms are governed by New York law. A European buyer will have to obtain that entire framework contractually.
Who owns the data?
The published documents settle less than they appear to. Nominal's privacy policy and terms of use both govern the marketing website only: form submissions, browsing data and cookies. They state that the site's content, graphics and logos belong to Nominal or its licensors, and that information sent through a form gives Nominal no access to the sender's systems. Nothing published says who owns the accounting data the platform processes, or what Nominal may do with it. That question lives in the customer agreement, which is not public, so a prospect has to obtain it directly before connecting a ledger.
Reuse rights
For the website, the rules are explicit. Nominal collects names, business emails, job titles, company details and the ERP a prospect uses, along with IP address, browser and pages visited. It uses them to answer demo requests, tailor demonstrations, improve its site and marketing, and stay in touch, and it shares them with CRM and marketing platforms, service providers, and legal authorities when required. It states plainly that it does not sell personal information. Website content may not be copied, reproduced or redistributed without written permission. Whether a customer's own accounting data may be reused, and in particular whether it may serve to train models, is nowhere addressed in the published documents.
Data retention & training
Hosting summary
Nothing is published about where data is hosted. The security page commits only to a cloud environment customers can trust: no country, no region, no cloud provider and no data residency option appears anywhere on the site, and a search across every page collected returned nothing further. The company is based in New York and its terms of use are governed by the law of the State of New York, which fixes the jurisdiction of the contract rather than the location of the data. The marketing site itself resolves to an Amazon-hosted anycast address in the United States, but that describes the website, not the platform where accounting data is processed, and the two should not be confused. There is no data processing agreement, no subprocessor list and no mention of international transfers. Any buyer with a residency requirement will have to ask directly.
Things to keep in mind
Risks and trade-offs to weigh before adopting Nominal.
- Total pricing opacity: the whole cost discussion happens in a sales conversation, with no public benchmark to negotiate against
- The published privacy policy covers the marketing site only, so the terms governing your accounting data are not public
- Not a word about the GDPR anywhere on the site, for a product that processes company financial data
- No hosting country, no data processing agreement and no subprocessor list are published
- Nothing states whether customer data may be used to train models, or whether that use can be refused
- Automation dulls scrutiny: when agents reconcile continuously, teams stop checking what they once checked by hand, and an audit trail only helps if someone reads it
- Impact figures such as 3x efficiency, 65% less hiring and 50,000 hours saved come with no methodology, and the SOC claims come with no report to verify them
Setup & Integrations
Technical difficulty
Low effort on paper, and vendor-led in practice. Nominal connects through an API, an on-premise deployment or data connectors, replicates transactions, balances and journal entries, and claims to require no ERP change, no custom script and no migration. One customer describes granting access to the accounting database and letting the team take it from there; another reports QuickBooks data arriving ten minutes after logging in. Nothing is self-service, however, and no public technical documentation lets you assess the work on your own side.
Deployment
Integrations
Behind Nominal
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement Nominal.
Frequently asked questions
What does Nominal actually do?
Do I have to change my ERP?
How much does it cost?
Can I try it without talking to sales?
Do the agents act on their own?
What security guarantees are claimed?
Is the GDPR covered?
Where is the data hosted?
Who is behind Nominal?
Should you pick Nominal?
Nominal is one of the more substantial entries in the agentic finance category, for a simple reason: the site shows its product instead of announcing it. Five agent families are documented individually, six ERP routes have their own pages, the mechanics of a bank reconciliation are described step by step, customers are named with role and company, and two funding rounds are dated and quantified with institutional investors behind them. A company address, two dated legal pages and two contact addresses are published. None of this proves the agents perform as claimed, but it is more than most tools in this space put on the table.
The gaps sit in one place, and it is a sensitive one for a product that touches a general ledger. The privacy policy and terms of use govern the marketing website, not the service: nothing published says who owns the accounting data the platform processes, how long it is kept, where it is hosted, whether subprocessors are involved, or whether it may train models. The GDPR is never mentioned. SOC 1 Type 2 and SOC 2 are claimed on a marketing page, with no report to consult. Pricing is absent entirely.
For a controller at a multi-entity company whose close runs on spreadsheets, the demo is worth the hour. Go in with a list: the service agreement and its data clauses, the SOC reports, the hosting region, the position on model training, the scope of access requested on the accounting database, and a price with the assumptions behind it.
- Choosing a selection results in a full page refresh.
- Opens in a new window.