INTELLIAGENT FOR SAS

Three AI Agents.
One SAS Investment, Fully Realized.

Three agentic AI products built natively on SAS® — so the platform you've already invested in does the reading, the reporting, and the reasoning your teams still do by hand.

SAS, SAS RAM, SAS Intelligent Decisioning, and SAS Visual Analytics are trademarks of SAS Institute Inc.

THE SUITE

Three Jobs Your SAS Platform Can Now Do on Its Own

Each agent takes over one of the three things your teams still do by hand.

The Reading

Compliance

Reads any document end to end and extracts every rule inside it — thresholds, conditions, approvals — then turns them into decisions that actually run.

The Reporting

RBC2 Reporting

Takes the quarterly capital submission from raw entity data to a governed, regulator-ready package, with every figure traceable to its source.

The Reasoning

IntelliAgent

Orchestrates across every SAS Viya component — planning the work, calling each service in turn, and carrying governed context from raw data through to finished dashboard.

HOW IT PLUGS IN

No New Stack. Your SAS Platform, Doing More.

Every agent runs on components you already license and operate. Select one below to see exactly what it touches.

Your SAS® Environment

SAS® RAM SAS® Intelligent Decisioning SAS® Studio SAS® Visual Analytics SAS® Model Manager SAS® Visual Investigator Built Decisions Your Governed Data

IntelliAgent spans the whole environment — reading, deciding, modelling, investigating, and reporting across every component you run.

Compliance · DOCUMENT INTELLIGENCE

From Document to Decision, Fully Automated

Autonomous AI agents that turn any document into live, explainable decisions.

Every organization runs on documents — policies, contracts, regulations, guidelines. Hundreds of pages someone has to read, interpret, and turn into rules by hand. It's slow, inconsistent, and the moment a document changes, the work starts over. So we asked: what if an AI agent could do that thinking for you, end to end?

ResultRule extraction that once took a specialist days now happens in seconds, straight from the source document.

Watch the Compliance Demo

Any Document

A policy, contract, or regulation. No template, no tagging.

Rules Extracted

Thresholds, conditions, approvals, and the logic behind them.

Decisions Run

Live flows score every record in a single pass.

Dashboards Out

Risk-rated, reasoned, and audit-ready for anyone to act on.

Challenge

Every organization runs on documents — policies, contracts, regulations, guidelines. Turning hundreds of pages into working rules is slow, manual, and inconsistent, and the moment a document changes the work starts over. Teams spend more time reading, interpreting, and re-keying logic by hand than acting on it — delaying decisions, driving up cost, and leaving compliance gaps every time the rules shift.

Solution

An AI agent inside SAS RAM reads any document, from any domain, the way an expert would — reasoning through the language and extracting every rule on its own: thresholds, conditions, approvals, and underlying logic. Those rules flow straight into SAS Intelligent Decisioning as live, executable decision flows, with no manual coding or re-keying. Data is then scored against them in a single pass, and a second AI agent triages every result — rating risk and recommending the next step — while SAS Visual Analytics turns it into audit-ready dashboards.

Days to Seconds

Rule extraction that once took a specialist days now happens in seconds, straight from the source document.

Zero Manual Coding

Rules become live, validated decision flows automatically — no re-keying, no hand-built logic, no waiting on IT.

Nothing Slips Through

Thousands of records scored in a single pass, with every exception surfaced automatically.

Fully Explainable

Every result is risk-rated and reasoned by AI, delivered as dashboards and an audit-ready report anyone can act on.

RBC2 · REGULATORY REPORTING

RBC2 Reporting — Automated, Auditable, Done Within Minutes

From days of spreadsheets to minutes of governed, regulator-ready capital submissions.

Built for Group actuaries managing multi-entity RBC2 submissions — auto-maps source tables, resolves multiple formulas through a full dependency graph, and generates audit-ready disclosure notes automatically. Every reconciliation, stress test, and board narrative runs through a single governed workflow — with maker-checker gates, full lineage on every figure, and zero manual handoffs between steps. The CFO dashboard is ready before the board meeting starts. The regulator submission is packaged in each authority's own format.

ResultTen business days recovered, every single quarter.

Watch the RBC2 Demo
The quarter close, by hand ≈ 10 business days
Reconcile entities Sequence formulas Rebuild the audit story Package
The quarter close, with RBC2 Minutes
Upload → resolve → reconcile → package

Challenge

Every quarter, the finance team spent multiple days manually reconciling entities' worth of data, sequencing multiple interdependent RBC2 formulas across spreadsheets and SAS scripts — where one broken cell derailed everything. With no audit trail, explaining any figure meant rebuilding logic from memory, under deadline pressure, every single time.

Solution

A single upload auto-maps all source tables and validates in real time. The engine resolves all multiple formulas through their full dependency graph — every figure audit-linked to its source. Reconciliation runs in the background, stress scenarios run overnight, and the final submission is packaged per regulator before the board meeting starts.

Days → Minutes

Per entity, every quarter, without exception.

Time Recovered

For judgment, not assembly.

4 Gates Kept, Zero Bottlenecks

Every sign-off stays. Only the waiting disappears.

IntelliAgent · AGENTIC ORCHESTRATION

One Agentic Layer, Orchestrating All of SAS® Viya

Not a conversational copilot — an agentic layer that orchestrates across the SAS® Viya platform, coordinating data, models, decisions, and analytics as a single governed workflow.

Work in SAS Viya rarely lives in one place: the data sits in CAS, the models in Model Manager, the logic in Intelligent Decisioning, the output in Visual Analytics. IntelliAgent sits above all of it — planning a task, invoking each component in the right order, and carrying state, context, and lineage across every hop, so an outcome that used to mean driving four tools by hand runs as one job.

ResultWork that used to span four SAS components by hand now runs as one orchestrated job.

Watch the IntelliAgent Demo

Challenge

A single outcome rarely lives in a single SAS component. Getting from raw data to a decision means moving by hand between CAS, Model Manager, Intelligent Decisioning, and Visual Analytics — re-keying context at each boundary, losing lineage between steps, and waiting on whoever owns the next stage. The work isn't hard; the coordination is.

Solution

IntelliAgent is an agentic orchestration layer over SAS Viya. It plans multi-step work, invokes SAS Studio, Model Manager, Intelligent Decisioning, Visual Investigator, and Visual Analytics in whatever order the task requires, and passes state and lineage between them automatically — all inside your existing SAS identity and permissions. Nothing is bolted on: governance travels with every step.

One Job, Not Four

A task that spanned four SAS components now runs end to end as a single orchestrated flow.

No Manual Hand-Offs

State and lineage carry across every component, so nothing is re-keyed between steps.

Governed by Design

Every step respects existing SAS roles, permissions, and data access — nothing bypasses governance.

DEPLOYMENT & GOVERNANCE

Nothing Leaves Your Environment

These are agents inside your SAS estate, not a service you send data to. Your controls stay exactly where they are.

Runs Where Your SAS Runs

On-premise, private cloud, or your existing SAS Viya deployment. Documents and data are processed inside your own estate — nothing is shipped to an external service for scoring.

Your Roles, Your Permissions

Every agent authenticates as the user and inherits the SAS entitlements they already have. No parallel permission model to maintain, and no way to read past an existing access boundary.

Every Figure Traceable

Extracted rules, resolved formulas, and scored results all carry lineage back to the source document or table — so an auditor's question is answered by opening a record, not rebuilding logic.

Sign-Offs Stay Human

Maker-checker gates and review steps remain exactly where your process puts them. The agents remove the waiting between steps, not the people who approve them.

Ready to Put SAS RAM, Intelligent Decisioning, and Visual Analytics to Work?

Talk to our team — we'll map Compliance, RBC2 Reporting, or IntelliAgent to your existing SAS workflow.