CloviTek AICloviTekAI
Start a project

Deterministic Reasoning

CloviRules LIVE

RETE-based forward-chaining rule engine with per-tenant isolation and full audit trails. The auditable, explainable alternative to LLM-only reasoning for compliance, eligibility checks, and policy enforcement.

Request Access ← All Products
RULE 1 RULE 2 RULE 3 INFERENCE AUDIT TRAIL Rule 1 fired → Fact A Rule 3 fired → Decision APPROVE RETE algorithm • Tenant isolation • Full audit log

What it does

Deterministic, auditable, explainable inference

CloviRules is a production-grade expert system for use cases where hallucination is unacceptable. Compliance gates, eligibility checks, lead scoring, diagnostic decision trees, and policy enforcement all demand deterministic reasoning with a full audit trail. CloviRules delivers exactly that — RETE forward-chaining inference with per-tenant rule namespace isolation and a logged trace of every fired rule.

⚖️
Use Cases
Where determinism is mandatory
Compliance gates (GDPR, HIPAA) Eligibility checks Lead scoring & routing Diagnostic decision trees Policy enforcement Approval workflows
🔒
Per-Tenant Isolation
Namespace-scoped rule namespaces
Tenant ID scoping No cross-tenant rule access Separate rule sets per client Multi-tenant safe
📝
Full Audit Trail
Every inference logged with rule trace
Rule firing sequence Input fact snapshot Decision outcome Timestamp + tenant ID Compliance-ready logs
⚙️
Technology Stack
Experta + business_rules + FastAPI
Experta 1.9.4 (RETE) business_rules (MIT) FastAPI :9102 Python 3.12 JSON/YAML rule DSL

Why CloviRules

The explainable alternative to LLM-only reasoning

Large language models excel at generation and natural language understanding, but they hallucinate. When a decision must be auditable, explainable, and deterministic — compliance gates, eligibility checks, policy enforcement — a rule engine is the correct tool. CloviRules provides that determinism without sacrificing developer experience or API simplicity.

How we compare

CloviRules vs the alternatives

How an auditable rule engine compares with asking an LLM or burying logic in hand-written code.

CapabilityCloviRulesLLM-only reasoningHard-coded if/else
Deterministic, repeatable decisionsGuaranteedVariesYes
Full audit trail of fired rulesBuilt inOpaqueManual logging
Rules change without code changesBuilt inPrompt editsCode changes
Per-tenant rule isolationBuilt inSharedYou build it
Explainable to a reviewerYesHardPartial
Part of the CloviTek stackNativeNoNo

Comparison of common approaches, as of 2026. Approaches vary; details may change.

Integrations

Connects to your stack

CloviRules slots in wherever a decision has to be correct and explainable.

Fa
Facts inSend the facts of a case and get a decision with the rules that fired.Input
Au
Rule authoringWrite and update rules declaratively, with no redeploy.Authoring
Ap
REST APICall the engine from any language over HTTP.API
Lg
Audit logEvery decision keeps a trace of the rules that produced it.Audit
Fl
CloviFlowGate an automation on a rule decision inside CloviFlow.Automation
Gw
CloviGatewayProgrammatic, versioned access through the platform gateway.Gateway
Where it's going

Built to keep getting better

CloviRules ships continuously. Here's what's live today, what's landing next, and where the product is headed.

Shipped
RETE, isolation & audit
  • Forward-chaining RETE inference with declarative rule authoring
  • Per-tenant rule isolation
  • Full audit trail of every fired rule
On the roadmap
Scale
  • A visual rule editor
  • Hybrid rules-plus-model decisions
  • Per-tenant isolation hardening at scale
How it works

From facts to decision in four steps

Author rules, send facts, get an explainable decision.

1
Author
Write your rules
Define rules declaratively in a simple format — no code.
→
2
Send
Post the facts
Send the facts of a case to the REST API.
→
3
Infer
RETE evaluates
Forward-chaining inference fires the matching rules deterministically.
→
4
Get
Decision + trace
Receive the decision plus the exact rules that produced it.
Who it's for

Who CloviRules is for

Deterministic decisions for the people who can't accept a guess.

Backend developers

Backend developers

Logic without spaghetti

Move branching decisions out of code and into editable rules.

Maintainable
Compliance teams

Compliance teams

Explainable decisions

Every decision carries an audit trail you can show a reviewer.

Auditable
Product teams

Product teams

Change rules, not code

Update eligibility or policy without a redeploy.

Agile
Put to work

Put CloviRules to work

What teams reach for CloviRules to do.

Eligibility checks

Deterministic yes/no decisions with a full audit trail.

Compliance gates

Enforce policy where a guess is unacceptable.

Scoring

Transparent, repeatable lead or risk scoring you can explain.

Decision trees

Encode expert logic as auditable, editable rules.

FAQ

Common questions

What is CloviRules?
A deterministic, explainable rule engine — RETE forward-chaining with per-tenant isolation and full audit trails — for decisions that must be correct and auditable.
Why not just use an LLM?
For compliance, eligibility and policy decisions you need repeatable, explainable results with an audit trail — which a rule engine guarantees and an LLM does not.
Can rules change without code?
Yes. Rules are authored declaratively and can be updated without a redeploy.
Is it isolated per tenant?
Yes. Each tenant's rules live in their own namespace and never cross over.

Expert System Inference

Deterministic reasoning with full audit trails

CloviRules brings RETE-based forward-chaining inference to compliance-sensitive workflows. Request access to the CloviRules API.

Request Access