3-Layer Architecture

The compliance engine that never guesses

Inlet's 3-layer compliance pyramid combines deterministic rule evaluation, conditional logic trees, and AI reasoning to check every ingredient against every regulation in every market -- in under 200 milliseconds.

10,000+
Regulations Indexed
40,000+
Ingredients Tracked
6,302
Ingredient-Regulation Links
<200ms
Average Check Time

Three layers. One answer.

Each layer handles the checks it is best suited for. Deterministic rules catch 80% of issues instantly. Conditional logic handles context-dependent edge cases. AI reasons about the rest.

1

Deterministic Rules

~80% of checksConfidence: 0.95 - 1.0

Hard-coded regulatory logic that handles the vast majority of compliance checks with near-perfect accuracy. No AI needed, no latency, no ambiguity.

Banned substance detection (EU Annex II, FDA prohibited list)
Maximum concentration limits per ingredient per market
Required warning labels and mandatory disclosures
Allergen declaration requirements
Preservative system validation against permitted lists
2

Conditional Logic

~15% of checksConfidence: 0.75 - 0.95

Rule trees that evaluate context-dependent regulations. The same ingredient may be compliant or non-compliant depending on the product category, intended user, or formulation details.

Category-specific limits (e.g., leave-on vs. rinse-off sunscreen)
Age-group restrictions (children under 3, babies)
Sub-region variations (EU member state derogations)
Combination effects (ingredient A + ingredient B interactions)
Claim-dependent rules (SPF claims trigger specific requirements)
3

AI Reasoning (Navi)

~5% of checksConfidence: 0.5 - 0.9

For the long tail of ambiguous cases that no rule engine can fully codify, Navi uses Claude to reason about novel ingredients, complex claims, and edge cases with full regulatory context.

Novel ingredient safety assessments without established limits
Ambiguous marketing claim compliance evaluation
Cross-market regulatory interpretation conflicts
Emerging ingredient classifications (CBD, microbiome, peptides)
Natural/organic claim substantiation analysis

How a compliance check works

STEP 01

Submit SKU

Upload your product formulation with ingredients and target markets. Supports PDF, CSV, XLSX, and image uploads with AI parsing.

STEP 02

Match & Resolve

Ingredients are fuzzy-matched against 40,000+ master records. Each is linked to every applicable regulation across your selected markets.

STEP 03

Check Rules

The 3-layer engine evaluates each ingredient against each regulation. Banned substances, concentration limits, conditional rules, and AI analysis run in parallel.

STEP 04

Report & Act

Get a detailed compliance report with per-ingredient results, confidence scores, and specific remediation guidance. Export as PDF or generate submission dossiers.

Global market coverage

Every market has different regulatory bodies, different banned lists, different concentration limits, and different labeling requirements. Inlet tracks them all.

United States

FDA / MoCRA

European Union

EC 1223/2009

United Kingdom

UK MHRA

Canada

Health Canada

Malaysia

NPRA / ASEAN

Singapore

HSA / ASEAN

Additional markets are added quarterly based on customer demand.

Why a layered approach matters

Deterministic Accuracy

80% of compliance is black and white: an ingredient is either banned or it isn't. Rules engines handle this with 100% accuracy and zero hallucination risk.

Context-Aware Precision

A 1% concentration may be fine in a rinse-off shampoo but restricted in a leave-on eye cream for children. Conditional logic captures these nuances without over-engineering.

AI Where It Adds Value

Novel ingredients, ambiguous claims, and evolving regulations are where AI reasoning shines. By reserving AI for the 5% that truly needs it, we maximize accuracy and minimize cost.

Ready to automate compliance?

Upload your first product and get a compliance report in under 60 seconds. No credit card required for the trial.