Our Supplier Vetting Process

Overview

A parts database is only worth using if the data in it is accurate. For procurement professionals evaluating whether to trust a sourcing tool, "vetted" needs to mean something specific — not just a claim on the marketing page, but a defined process with real checkpoints.

Here's what vetting actually means at 3E Technology: every supplier, regardless of how they enter the database, clears two layers of review before publication. First, an automated verification system checks data accuracy, completeness, and flags inconsistencies. Then a human admin reviews the entry and makes the final approval decision. No supplier goes live without that human sign-off.

That two-layer standard applies to AI-discovered suppliers, self-submitted vendor registrations, and entries carried over from 3E's nearly 40-year sourcing history. The rules are the same across all entry paths.

The Two Layers of Review

Layer 1: Automated AI Verification

When a new supplier is identified — whether through AI research or vendor self-submission — an automated verification system reviews the record before any human sees it. This layer checks:

• Contact information validity: Are the phone numbers, email addresses, and physical addresses consistent and reachable? Does the domain exist? Are there obvious formatting errors or placeholder data?
• Data completeness: Does the entry have the minimum fields required to be useful to a buyer — company name, location, at least one contact method?
• Duplicate detection: Is this company already in the database under a different name or slight variation? Duplicate suppression prevents the same supplier from appearing multiple times with inconsistent data.
• Consistency cross-check: Do the company's listed brands and parts match what an AI verification pass can confirm about their actual catalog?

Records that clear this automated pass move to human review. Records that fail are flagged — either for correction before resubmission or for rejection if the data is fundamentally invalid.

Layer 2: Human Admin Final Review

Every record that clears automated verification reaches a human admin before publication. The admin review catches edge cases that automated systems miss:

• Legitimate-looking but defunct companies: A company can pass all automated checks — valid contact info, plausible history — while actually having closed or been acquired years ago. Human reviewers catch these with context that AI systems don't have.
• Subtle contact errors: A phone number that's technically valid but rings to a different business. An email domain that exists but belongs to a personal account, not the company. These require judgment.
• Bad actors: Fraudulent listings, companies misrepresenting their capabilities, or entries that raise red flags that don't fit a simple automated rule.

Admins can approve, reject, or flag entries for additional research. Only approved entries reach the live database.

Extra Scrutiny for Brokers and Surplus Dealers

The standard two-layer vetting process applies to all suppliers. Broker and surplus dealer entries receive additional scrutiny on top of that baseline.

The reason is documented risk:

42.75% of suspect counterfeit component reports in 2024 involved obsolete or end-of-life parts — exactly the inventory that independent brokers and surplus dealers carry.

— ERAI 2024 Annual Report

Counterfeits in the broker channel have caused component failures in aerospace, defense, medical device, and telecommunications applications.

This doesn't mean brokers are bad actors — most are legitimate businesses that play an essential role in sourcing parts that authorized distributors no longer carry. It means the vetting standard for broker entries is higher, because the consequences of a bad entry in that category are more severe.

For broker entries, the review process includes additional checks on business legitimacy, confirmation that they're a known entity in the industry, and closer scrutiny of any contact information that can't be independently verified.

What "Vetted" Covers — and What It Doesn't

The vetting process is designed to ensure that every entry in the database represents a real, reachable business that legitimately operates in parts supply. It covers:

• Data accuracy: Contact information is verified to the extent possible without placing a call to every supplier.
• Business legitimacy: The company is a real operating business, not a shell, defunct entity, or fraudulent listing.
• Data completeness: The entry has enough information to be useful — buyers need at least one way to contact the supplier.
• Duplicate suppression: The same company doesn't appear multiple times with conflicting data.

What vetting doesn't cover:

• Current inventory confirmation: Whether a specific supplier has a specific part in stock at this moment. Inventory changes daily. Buyers should confirm availability directly with the supplier.
• Pricing: 3E Technology doesn't track pricing. Buyers negotiate directly.
• Quality certification validation: Whether a supplier currently holds AS9100, ITAR, or other certifications. Certifications expire and lapse; buyers sourcing for regulated applications should verify certification status directly with the supplier.

Continuous Re-Verification of Existing Suppliers

Vetting isn't a one-time event. The same AI research system that discovers new suppliers also runs continuous re-verification of existing ones. When research agents investigate a search job, they cross-reference results against what's already in the database — catching cases where a company has moved, changed phone numbers, rebranded, or gone out of business.

User-flagged reports feed directly into this process. When a buyer reports that a contact is incorrect or a company appears defunct, that triggers a review of the entry. The admin team investigates, updates or deactivates the entry, and logs the change with a provenance record so there's a trail of what was found and when.

Over time, this continuous re-verification means the database gets more accurate, not less, as entries age. Stale data gets corrected or removed; confirmed data accumulates a track record of successful buyer contact.

Provenance Data on Every Entry

Every company in the 3E database carries provenance metadata: when the entry was added, how it was sourced (heritage sourcing relationship, AI-discovered, or vendor self-submitted), and when it was last verified. This data isn't visible to buyers in the search results, but it drives internal prioritization — entries that haven't been re-verified recently are flagged for review before entries with a recent confirmation.

For the heritage entries that form 3E's original database, provenance goes back to the actual sourcing transactions that created them: companies added because real buyers needed real parts and 3E found who had them. That transactional origin is a different kind of vetting than any automated system can replicate — it means the company supplied real parts in real procurement scenarios, not just that it passed a data-quality check.

What Makes This Different

Most parts databases don't describe their vetting process in specific terms because there isn't much to describe. Self-registration platforms publish what suppliers tell them about themselves. Aggregator databases pull from distributor feeds with no independent verification. Scraped databases reflect whatever was publicly indexable at the time of the scrape.

3E Technology's vetting process is different in three ways. First, every entry has a human sign-off — no supplier goes live from automated processing alone. Second, the two-layer process catches both data-quality problems (automated) and judgment-call problems (human review). Third, vetting is continuous, not a one-time check at entry — re-verification runs alongside new discovery.

The result is a database where "vetted" is a specific claim with a specific meaning: two layers of review, human approval required, and continuous re-verification over time.

Summary

Every supplier in the 3E Technology database cleared automated AI verification and human admin review before publication. Broker and surplus entries receive additional scrutiny given the documented counterfeit risk in that channel. Existing entries are continuously re-verified as the AI research system runs and as users report issues.

For buyers, this means a database where every contact is a real, reachable business — not a self-reported listing or an unverified scrape result. No markup, no middleman, best deal forever. Search the database and reach suppliers directly.