How does the NV3 free trial work?

Sign up for a free license key and get 500 complimentary transactions to test NV3 against your own data using the trial endpoint. No credit card and no obligation.

New — Name Validation 3 is now powered by machine learning

Verify, classify, and parse any name in one API call.

NV3 is Service Objects' machine-learning name-verification API — trained on billions of name pairs across 80 cultures. Validate, correct, classify, and match names in a single sub-second ValidateName request, with a 0–1 confidence score on every result.

Start Free — 500 Transactions See the API

No credit card required REST / JSON Sub-second responses

GET /NV3/ValidateName JSON

// request
{ "FullName": "Dr. John A. Smith Sr." }

// response
{
  "NameResult": {
    "IsValidName":   true,
    "Classification": "FullName",
    "Confidence":    0.98,
    "ParsedName": {
      "Prefix": "Dr.", "First": "John",
      "Middle": "A.", "Last": "Smith",
      "Suffix": "Sr."
    },
    "IsCommonFirstName": true,
    "Warnings": []
  }
}

Validity, classification, confidence & parsed parts — one round trip.

Trusted by data teams at brands that can't afford bad records

The engine, rebuilt

A machine-learning engine, not a rulebook.

The previous generation matched names against a fixed 40-step rulebook. NV3 replaces it with machine-learning models trained on billions of real name pairs — so it can infer, validate, and classify names that never appear in any lookup table.

Billions

of first- & last-name pairs in the training data

80 cultures

of names across the world's languages and scripts

0.00 – 1.00

continuous confidence score returned on every call

Everything one call returns

Six signals from a single request.

NV3 doesn't just say valid or invalid. Every ValidateName call returns parsing, classification, validity, confidence, matching, and risk — no separate services to stitch together.

Parse into components

Break any name into prefix, first, middle, last, and suffix — ready to map straight into your CRM fields.

Validate with confidence

Get a true/false validity flag plus a continuous 0–1 confidence score, so you can auto-accept, review, or reject by threshold.

Person vs. business

An ML classifier tells “Michael's Diamonds” (a business) from “Michael Diamond” (a person) — routing each record the right way.

Global & non-Latin

Automatically translates and transliterates non-Latin scripts — “박민준” comes back as “Park Min-Jun” — across 80 cultures.

Match & de-dupe

Similar-name, related-name, and duo-name outputs help you find and merge duplicate contacts before they multiply.

Risk & fraud warnings

Flags vulgar, bogus, celebrity, dictionary, garbage, and negative-sentiment names — even decoding leet-speak like “h3110” → “hello”.

Understands names like people do

It knows what it's looking at.

NV3 returns a Classification for every input — FullName, FirstName, LastName, DuoName, Business, Dictionary, Garbage, or Unknown — with the confidence behind it. Here's the same engine on four very different inputs.

Michael Diamond Person · FullName0.97

Michael's Diamonds Business0.95

박민준 Translated → Park Min-JunPerson · non-Latin transliteration

asdf qwerty Garbageflagged

Live in three steps

From sign-up to first validated name in minutes.

No SDK lock-in. If your stack can make an HTTPS GET request, it can call NV3.

Get your free key

Call ValidateName

Send a GET to /NV3/ValidateName with a full name or discrete parts and your AuthID.

Act on the result

Read IsValidName, Classification, Confidence, and warnings to accept, route, or reject automatically.

Read the API documentation

Why clean names pay off

Bad name data is expensive. NV3 stops it at the source.

Validating names at the point of capture protects revenue, productivity, and trust across the business.

$12.9M

average annual cost of poor data quality per organization

Gartner, 2020

27.3%

of a sales rep's time wasted on inaccurate B2B data

EBQ

40%

more revenue from personalization at faster-growing companies

McKinsey, 2021

$3.1T

estimated annual cost of bad data to U.S. business

IBM, 2016 (via HBR)

Figures are industry context from the cited sources — directional, not a guarantee of NV3-specific ROI.

Where teams put NV3 to work

Built for developers. Loved by data & ops leaders.

Fraud & fake-lead prevention

Block garbage, bogus, and vulgar names at point of entry to keep fake sign-ups out of downstream systems.

CRM & database hygiene

Correct misspellings, standardize casing and components, and flag duplicates via similar-name matching.

Personalized outreach

Correctly-cased, correctly-parsed names power salutations and campaigns without embarrassing errors.

KYC & onboarding

Person-vs-business classification and name standardization support onboarding and compliance workflows.

E-commerce & checkout

Screen names at checkout and on lead forms, complementing Lead, Email, Phone, and Address validation.

Deduplication & matching

Similar and related-name outputs merge duplicate contacts and keep reporting clean.

Enterprise reliability, out of the box

The infrastructure that has powered Service Objects' validation APIs since 2001 — and over 8 billion validated transactions.

99.999%

Financially-backed uptime SLA (~26 sec/month)

24/7/365

Support with dedicated engineers

8 billion+

Transactions validated to date

500

Free transactions to start

Already on NV2?

What changed in version 3.

NV3 is the machine-learning successor to Name Validation 2. NV2 remains available — but here's why teams upgrade. (NV3 is REST/JSON only.)

CapabilityName Validation 2Name Validation 3

Core engineRules-based “40-step” processMachine-learning models

Reference data~1.4M first + 2.75M last namesBillions of name pairs

Culture coverageGlobal DB, accented characters80 cultures & languages

Confidence scoringInteger scores 0–5Continuous 0–1 score

ClassificationPerson / Business / Unknown8 classes incl. Business, DuoName, Dictionary, Garbage

Non-Latin scriptsAccented charactersAuto translate / transliterate

Leet-speak & sentiment—Leet decode + negative-sentiment flags

ProtocolsREST & SOAP · JSON & XMLREST · JSON

Built for developers

One endpoint.
Failover built in.

A single ValidateName GET returns JSON. Reference client libraries handle automatic multi-datacenter failover, with sample code for C#, Python, and Node.js.

Automatic failover — live calls retry the backup data center on error/timeout (15s default).
Simple auth — pass your AuthID license key as a query parameter.
Batch & CRM — real-time API, the DataTumbler desktop app, NuGet, Dynamics 365 & Salesforce.

Primary https://sws.serviceobjects.com/NV3/ValidateName

Backup https://swsbackup.serviceobjects.com/NV3/ValidateName

Trial https://trial.serviceobjects.com/NV3/ValidateName

Get your free key View developer guide

quickstart.shcURL

# One GET request validates a name
curl "https://sws.serviceobjects.com/NV3/ValidateName?\
FullName=Dr.%20John%20A.%20Smith%20Sr.&AuthID=YOUR_KEY"

# → 200 OK
{
  "NameResult": {
    "IsValidName": true,
    "Classification": "FullName",
    "Confidence": 0.98,
    "ParsedName": { "First": "John", "Last": "Smith" },
    "Warnings": []
  }
}

Illustrative — see the docs for the full response schema.

Secure & compliant

Your data stays yours. We never resell it.

NV3 runs on the same certified, privacy-first, bank-grade-secure platform trusted by regulated industries.

SOC 2 GDPR CCPA / CPRA Bank-grade SSL

Questions, answered

Frequently asked questions

How does the free trial work?

Sign up for a free license key and get 500 complimentary transactions to test NV3 against your own data. Point them at the trial endpoint — no credit card, no obligation.

What makes NV3 different from NV2?

NV3 replaces the rules-based “40-step” engine with machine-learning models trained on billions of name pairs across 80 cultures. It adds person-vs-business classification, a continuous 0–1 confidence score, non-Latin transliteration, leet-speak decoding, and negative-sentiment flags. NV2 remains available; NV3 is REST/JSON only.

What does a ValidateName call return?

IsValidName, a Classification (FullName, FirstName, LastName, DuoName, Business, Dictionary, Garbage, or Unknown), a 0–1 Confidence, the ParsedName components, similar/related names for matching, plus Notes and Warnings (vulgar, bogus, celebrity, business, dictionary, negative-sentiment, garbage).

How do I integrate it, and is there failover?

It's a single REST GET returning JSON, authenticated with your AuthID. In live mode the reference client libraries automatically retry the backup data center if the primary errors or times out (15-second default). Sample code is provided for C#, Python, and Node.js.

How much does it cost?

Start free with 500 transactions, then move to simple volume-based pricing sized to your usage. Contact Service Objects for a quote for your volume.

Is my data kept private?

Yes. All traffic is encrypted, and the platform is SOC 2 certified and GDPR / CCPA-CPRA aligned. Service Objects does not resell or share the data you submit.

Put the ML name engine to work — free.

Get 500 transactions on the house and run NV3 against your own data. No credit card. No commitment.

Start Free — 500 Transactions Talk to an expert

Live in minutes 99.999% uptime SLA 24/7/365 support

Verify, classify, and parse any name in one API call.

A machine-learning engine, not a rulebook.