# SAMI — Conversational Security Analytics

Source: https://s3msecurity.com/products/sami-conversational-security-analytics/
Last updated: 2026-11-03

SAMI is a conversational security analytics layer that turns NAC and cellular Zero Trust telemetry from ConnGuard and APNZone into plain-language answers. It provides SIM-level insight and anomaly flags with no new agents or hardware, and is read-only by design: it never changes policy or sessions, and points to APNZone when action is needed.

Ask your network anything. SAMI turns your NAC and cellular Zero Trust telemetry into plain-language answers.

## Key capabilities

### Ask in your own words

Type a question the way you'd ask a colleague. No query language, no saved filters, no dashboard-building.

### Source-backed answers

SAMI reads live NAC and cellular telemetry and returns a source-backed answer — ranked lists, anomalies, SIM and device detail.

### SIM-level insight

Reasons over the subscriber (IMSI / SUPI), device hardware (IMEISV / PEI), cell and region location (ULI) and access technology (RAT), identifiers that don't exist in Wi-Fi-only tools.

### Detect & act in one loop

SAMI surfaces the risk; the same platform enforces it. Analytics and control share one system — no export, no swivel-chair.

### No autonomous actions

SAMI never quarantines, disconnects, or rewrites policy. It tells you what's happening; you decide what happens next.

### Tenant-isolated

Each environment's data stays its own. Nothing is pooled across customers — built for sovereign and regulated deployments. Where model improvements are shared, raw subscriber data never is.

### Audit-friendly

Every answer is source-backed and traceable to the telemetry behind it — clear provenance for compliance and review.

## How it works

1. **Ask — Say it your way** — Type a question the way you'd ask a colleague. No query language, no saved filters, no dashboard-building.
2. **See — Get a precise answer** — SAMI reads live NAC and cellular telemetry and returns a source-backed answer — ranked lists, anomalies, SIM and device detail.
3. **Know — Decide what's next** — Understand what's happening and why it matters. When action is needed, SAMI points you to APNZone — your team stays in control.

Read-only by design • No new agents • Built on data you already own

## Your access data has the answers. Getting them takes a ticket and three days.

ConnGuard and APNZone already capture who connected, on what device, from where, over which technology. But asking a real question — “which SIM is acting strange?” — still means dashboards, filters, exports, and someone fluent in the query language.

### Locked in dashboards

The insight exists, but it lives behind saved views and filters only specialists can build. Most people never get to ask.

### Wi-Fi-shaped tools

Existing network assistants answer Wi-Fi troubleshooting questions. None of them speak SIM, IMEI, or cellular location.

### Slow to answer

By the time the report is pulled and read, the anomaly has already moved — a different cell, a different device, a different SIM.

## Ask. See. Know.

SAMI sits on top of ConnGuard and APNZone as a read-only analytics layer. You ask in plain language; SAMI reads your live telemetry and answers — with the SIM-level detail only the cellular layer holds.

## One assistant for every access question.

From “how many users are online?” to “which SIM is out of region?” — SAMI answers across your wired NAC and cellular Zero Trust layer, in the same plain language.

### Example: cellular question

**Question:** Any SIM connecting from outside its authorized region right now?

**Answer:** 1 session outside its authorized region.

| Field | Value |
| --- | --- |
| Flagged SIM | IMSI 310 410 ···· 4471 |
| Device | EdgeBoX-EB12 |
| Authorized | HQ-Field · 0x1A12 |
| Current | Out-of-perimeter · 0x7F90 |

Recommended: quarantine in APNZone — SAMI won't apply this itself.

### Example: network access question

**Question:** Top 5 users by traffic today (live preview: 4,210 SIMs · 9 sites, read-only)

| User | Traffic today |
| --- | --- |
| a.kovac | 84.2 GB |
| m.silva | 61.7 GB |
| guest-7741 | 48.3 GB |
| r.okafor | 39.1 GB |
| k.tan | 27.8 GB |

*Interactive preview · sample data, masked for privacy.*

### Questions SAMI answers

**Network access · NAC**

- Users online today
- Failed password attempts
- NAS device count
- Endpoints failing posture
- Active guest sessions

**Cellular Zero Trust**

- Sessions outside the region
- Hardware changes (24h)
- NR → 2G downgrades
- Dormant SIMs reactivated

## Answers grounded in the layer no one else sees.

Wi-Fi assistants stop at the access point. SAMI sees the SIM.

Because SAMI reads ConnGuard and APNZone data, it reasons over identifiers that simply don't exist in Wi-Fi-only tools — the subscriber, the hardware, the cell, the access technology. That's what lets it answer questions no other assistant can.

| Identifier | Field |
| --- | --- |
| Subscriber identity | IMSI / SUPI |
| Device hardware | IMEISV / PEI — binding |
| Location | Cell / region (ULI) — geofence |
| Access technology | RAT — downgrade detection |

## Read-only by design.

SAMI answers, ranks, and flags — it never changes policy or sessions. Recommendations point to APNZone, where your team acts. That keeps SAMI safe to give to analysts, auditors, and executives alike.

SAMI tells you what's happening. You decide what happens next.

## Built for environments where one stray SIM is a real incident.

- **Defense & Military:** Field & tactical SIM fleets
- **Law Enforcement:** Patrol, body-cam, fleet
- **Energy Sector:** OT, SCADA, grid telemetry
- **Telecom & ISPs:** Carrier-scale RADIUS
- **Healthcare:** Clinical & medical IoT

## Put SAMI on your network.

In a 30-minute demo, we'll show SAMI answering real questions about an environment like yours — SIM-level visibility, anomaly detection, and the read-only model your security team can trust.

## Frequently asked questions

### Is SAMI an autonomous agent?

No. SAMI is read-only by design. It analyzes your data and answers questions — it never changes policy, quarantines devices, or terminates sessions. When action is warranted, SAMI recommends it and links you to APNZone, where your team applies the change.

### What data does SAMI use?

Your existing NAC and cellular Zero Trust telemetry from ConnGuard and APNZone — authentication events and accounting records you already collect. SAMI adds an analytics layer on top; it doesn't introduce new collection.

### Does SAMI need new agents or hardware?

No new endpoint agents and no new hardware. SAMI reads the telemetry your platform already produces, so there's nothing to deploy to devices.

### Can SAMI really answer cellular questions other tools can't?

Yes. SAMI reasons over SIM (IMSI), hardware (IMEISV), cellular location, serving network, and access technology — fields that only exist on the cellular layer S3M operates on. Wi-Fi-oriented assistants don't have that data to answer from.

### How does SAMI fit with ConnGuard and APNZone?

SAMI is the conversational analytics layer on top of the platform. ConnGuard and APNZone enforce access and capture the data; SAMI makes that data answer questions in plain language and points back to APNZone when it's time to act.
