We eliminate software.

The integrations, scripts, workflows, and internal tools your operations run on. Each one began as a business decision and became a software project to specify, build, deploy, and maintain forever.

Inferal runs your business rules directly. Define what should happen and when. Inferal connects to your data and acts on it continuously, with no code to build or maintain.

Position limits

−0 +0

src/main/java/…/integration/OmsBridgeClient.java−4,733
src/main/java/…/exposurecalculator/*.java−15,063
src/main/java/…/pretradegate/*.java−17,426
src/main/resources/templates/limit-overrides/*.html−29,072
src/test/java/…/PositionLimitsIT.java−1,099
pom.xml−544

Stack

Java
Spring Boot
Oracle
Kafka
Kubernetes

Lines eliminated: −0; –
Maintenance cost saved: $0; AI spend $0
Dev time saved: 0dev-months; –

Late-delivery response

−0 +0

src/main/java/…/integration/CarrierSyncClient.java−4,165
src/main/java/…/etarecalculation/*.java−2,101
src/main/java/…/consigneenotify/*.java−1,503
k8s/late-delivery-sweep-cronjob.yaml−393
src/main/resources/templates/exceptions-panel/*.html−5,487
src/test/java/…/LateDeliveryIT.java−1,265
pom.xml−315

Stack

Java
Spring Boot
Oracle
Kafka
Kubernetes

Lines eliminated: −0; –
Maintenance cost saved: $0; AI spend $0
Dev time saved: 0dev-months; –

The manifesto

We have mistaken translation work for software engineering.

The translation problem

AI cuts the cost of writing code. The software lifecycle remains.

Language models can shorten the path to a first implementation. But the code still becomes another source of truth to review, deploy, secure, observe, update, and keep aligned with changing business intent.

Translation≈ $0Lastshours
Someone decides what should happen, under which conditions, and why it matters to the business.
Business contextIntent explicit
Capacity≈ $470/ five-person hour¹Lastsdays
People explain, challenge, and interpret the decision together. Shared context becomes conversation and notes.
Shared discussionInterpretation begins
Capacity≈ $25–$50/ ticket¹Lastsa few hours
The agreement is compressed into a unit of work: a request, acceptance criteria, and a place in the backlog.
Work itemIntent summarized
Capacity≈ $3.8K/ developer-week¹Lastsmonths
The request becomes precise enough to implement. Ambiguity narrows, but business rationale starts to fall away.
Technical contractBehavior defined
Capacity≈ $16K/ developer-month¹Lastsmonths
```
if (condition.matches(context)) {
  await execute(requiredAction);
}
```
Application logicMeaning encoded
Capacity≈ $49K/ developer-quarter¹Lastsyears
The implementation enters a specific environment with configuration, dependencies, versions, and release constraints.
Runtime environmentBehavior live
Cost≈ $59K–$85K/ developer-year²Lastsuntil retired
Software keeps accruing obligations after release. Every change must preserve business intent while the technical world beneath it keeps moving.
Ongoing systemCost compounds
- Business alignment
  Keep behavior synchronized with changing policies, workflows, priorities, and operating reality.
- Security and CVEs
  Monitor dependencies, privileges, and attack surfaces, then assess and remediate new vulnerabilities.
- Updates and compatibility
  Absorb runtime, framework, platform, integration, and data migration changes without breaking behavior.
- Software lifecycle
  Software must be operated, observed, upgraded, migrated, and eventually retired. Each stage adds dependencies, coordination, and risk that outlive the original decision.

Six handoffs. Then maintenance. One decision, with translation cost compounding at every step.

How it works

Make the rule the source of truth.

This is what data-native means: rules, relationships, and constraints remain explicit and connected to the live facts they govern. Inferal evaluates them continuously and sends the resulting actions to the systems and agents that carry them out.

The people who know the work can finally speak for it.

Operational teams

State what should happen.

Define conditions, exceptions, and desired outcomes in terms that remain connected to the work.

Risk and governance

Keep intent inspectable.

Inspect what governs a decision and trace the facts and rules behind every action.

Engineering

Build the foundations.

Build the algorithms and core capabilities unique to the business. Operational teams own the changing rules around them.

The result is less software between a decision and its execution.

Operational policyLate-delivery response

Delivery exception

When a shipment will arrive after its committed time, open an exception and notify the consignee.

Live facts

Shipment is manifested
Vehicle location is current
Estimated arrival is later than committed time

Result

Late-delivery exception created
Consignee notification requested
Follow-up scheduled as another rule

This is an operational model.

The rules, facts, relationships, and policies you just saw form a live, governed representation of how the business works. Ontologies give that model shared meaning across systems and agents.

Explore operational models

Decision provenance

Every outcome keeps its why.

Because the rule stays connected to the data it evaluated, every outcome retains the exact instruction, evidence, and moment behind it.

Decision receipt

Record 0142

Delivery exception

A late delivery was caught while there was still time to respond.

What changed

A delivery is now expected late.

The vehicle’s latest position moved the expected arrival from 16:42 to 17:18.

What should happen

Late deliveries require intervention.

If the expected arrival is later than 17:00, open an exception and notify the consignee.

What happened

The exception was opened.

The consignee notification was requested automatically.

Recorded at: 14:32:08, when arrival changed to 17:18
Policy in force: Late-delivery response, effective 12 June
Evidence at that moment: Vehicle position at 14:32, arrival 17:18, delivery due 17:00

Customer perspectives

What people see in Inferal.

“In an era where source code became a commodity, being able to get exactly the data you want in real-time is the moat you need to develop powerful agents. Inferal Relay provides this moat for us.”
Krisztián SzabóCTO of MantaSee Inferal Relay

“If the data is the source of truth, then business decisions should be made as close to it as possible. Inferal lets you push your business rules right to the data itself.”
Nisan HaramatiCo-Founder at Graphium Labs

See what others say

“We monitor DNA sequences, protein data, antibody assays, and need to act on new relationships continuously. Inferal executes rules close to the database, giving us a fast way to alert AI agents and researchers.”
Morten KroghCTO at Bionamic

“Clearly the future for rule-based agentic behavior for all verticals. Everyone else is using agents in a reactive way and Inferal is using them proactively.”
Sean MaginSoftware Engineer

“What stands out about Inferal is how directly their approach translates into faster execution and lower complexity.”
Michel PelletierFounder at OneSparse

“The database sees everything going in and out, but it usually just stores it. Inferal turns those changes into instant actions without adding a bunch of extra systems.”
Rob PankowCEO at simplyblock

Writing

From the Inferal journal

View all writing ↗

Start with one process

Which business process is waiting on engineering?

Start with one that changes too often, crosses too many systems, or lives in too much code.

Run your business, not software.

Unblock it