⚡ Infrastructure Products · Two products, one engineer

Software for the people
who run the systems.

Two purpose-built infrastructure products, engineered and deployed into your environment by a systems engineer with 20+ years in real enterprise networks. Server Guardian gives your team eyes on the network. ATLAS gives them a private AI assistant that answers only from your own knowledge — on hardware you control. Neither is sold in an app store; both are acquired by talking to us directly.

🛰 Real-Time Monitoring

Server Guardian

See your network. Actually.

A real-time operations dashboard for IT teams — host and service health, group fleet views, performance trending, alert management, and service drill-down. It reads from the monitoring system you already run, so nothing gets ripped out. Built by a systems engineer, deployed by a systems engineer, into your environment.

Platform Overview
Server Guardian real-time overview dashboard with health gauge
OverviewHealth gauge · problem grid
Server Guardian host groups fleet view
HostsGroup view · service rollup
Server Guardian group health report with per-server health rings and CPU, memory, disk meters
Group Health ReportHealth ring per server · CPU / Memory / Disk / Net I/O · service matrix · top alerts
Server Guardian services list with state filter chips
ServicesState filters · plugin output
Server Guardian performance trending charts over time
Trending60 min → 30 days · multi-server
How It Works
1

Your monitoring system

Nagios XI, InfluxDB, Zabbix, Fortinet, UniFi — any API-accessible source. We connect to what you have. No migration.

2

Backend connector

A lightweight FastAPI service we deploy in your environment during the remote wiring session. Aggregates, caches, and secures the pipeline.

3

Server Guardian app

The dashboard your team opens — Android, iOS, or Windows desktop, compiled for your platform and distributed directly to your people.

What's Inside
  • Real-time dashboard — health score, hosts up/down, critical services, storage volumes
  • Host & group views — searchable, color-coded by health, service rollup per host
  • Group health reports — per-server rings, CPU/Memory/Disk/Page File/Net I/O, service matrix
  • Performance trending — sparkline charts, Last 60 min → 30 days, multi-server overlay
  • Alert management — push notifications, per-category sounds & voice, quiet hours
  • Cross-platform & white-label — one codebase, your branding on request
No rip-and-replace. Server Guardian is a presentation layer over your existing monitoring — no new agents on your servers, no data migration. It inherits whatever coverage you already have.
Remote installation. Deployment is handled remotely over your VPN or a secure remote-access method you choose. No hardware to ship, no travel. You provide the backend host (VM or physical) and remote connectivity; the server and hypervisor are not included.
Pricing
Software licensePerpetual · specified platform(s)
$1,495 +
Professional implementationScoped per environment · one-time
$3,000–$8,000
Typical complete engagementSoftware + implementation
$4,500–$9,500

Annual maintenance & support optional at 20% of license value/year (security patches, compatibility updates, priority support). Perpetual license — you own what you buy.

🧠 On-Premise AI Assistant

ATLAS

Your knowledge. Your AI. Your walls.

A private, self-hosted AI assistant that answers only from your documents — never the public internet, never a third-party cloud. ATLAS pairs a self-hosted language model with retrieval-augmented generation (RAG) and enforces your Active Directory / LDAP permissions, so each person only gets answers from the folders their access allows. It ships as a polished mobile app and an embeddable web chat widget for any internal site. Everything runs inside your network, on hardware you own.

Platform Overview
ATLAS AI assistant home screen with knowledge folder chips and greeting
WorkspaceFolder-scoped · voice or text
ATLAS answering a question grounded only in the organization's own documents
Grounded AnswersFrom your docs, not the web
ATLAS embeddable web chat widget open on an internal site
Embeddable Web WidgetOne script tag drops a per-user, access-scoped assistant into any internal site
ATLAS directory sign-in screen
Directory Sign-InAD / LDAP · per-user RBAC
ATLAS server settings pointing the app at your own AI server
Your ServerPoints at your on-prem AI
How It Works
1

Your GPU AI server

A self-hosted language model runs on your GPU hardware inside your network. Your documents are indexed into a private knowledge base. Nothing leaves your walls.

2

RBAC & retrieval layer

Each request carries the signed-in user's directory token. Your Active Directory / LDAP groups decide which knowledge folders answer — enforced server-side, with an audit trail.

3

ATLAS app + web widget

Your people ask by voice or text from the mobile app, or from the chat widget embedded on your intranet. Answers are grounded in your documents and cited to the folders they came from.

What's Inside
  • Answers only from your knowledge — retrieval-augmented, grounded in your documents, not the open web
  • Per-person access control — AD/LDAP RBAC decides what each user's assistant can see
  • Voice & text — speak a question or type it; natural, in-house assistant experience
  • Mobile app + web widget — a full app for your team and a one-tag embeddable widget for any site
  • Configurable & white-label — point it at your own model/server; your name, accent, and branding
  • Fully private & on-prem — no third-party AI cloud; data stays inside your network
Dedicated GPU hardware is required — and is NOT part of this purchase. A self-hosted AI model needs one or more dedicated enterprise GPUs (e.g., NVIDIA A- or L-series) with sufficient VRAM, plus a suitable server to host them. You provide and own that hardware. We will help you spec it, but we do not sell, lease, host, or supply it.
We provide the software and the implementation — not the infrastructure. This engagement does not include GPUs, servers, storage, virtualization, networking, model licensing, or any cloud/compute. You supply the AI host and all supporting infrastructure; ATLAS is deployed onto it. Bayou Byte Studio is not responsible for procuring, licensing, or supporting the underlying infrastructure the platform runs on.
100% remote deployment. All work is performed remotely. You provide secure remote access — VPN, RDP, SSH, or an equivalent method of your choosing — for our engineer to install, wire, and verify the platform. No on-site visit is included.
Pricing
Software licensePerpetual · ATLAS app + web widget
$28,000 +
Professional implementationOn-prem LLM + RAG + RBAC · scoped
$65,000–$135,000
Typical complete engagementSoftware + AI implementation
$95,000–$165,000

Reflects a fully on-premise, privately hosted, access-governed AI platform. Comparable enterprise on-prem AI/RAG deployments run $80,000–$150,000+; ATLAS is delivered as a productized platform with a mobile app and embeddable widget. Excludes all hardware and infrastructure (see above). Annual maintenance & support optional at 20% of license value/year — model updates, knowledge re-indexing, security patches, and priority support.


Why Bayou Byte Studio

Built by engineers. Deployed by engineers.

You're not buying a subscription to another SaaS dashboard or a black-box AI service that ships your data off to someone else's cloud. Both products are engineered and deployed into your environment by someone who has done this work in real enterprise networks — with your data staying under your control.

Your data stays yours

Server Guardian reads from systems you already run. ATLAS runs entirely on your own hardware. Nothing is shipped to a third-party cloud.

Professional implementation included

Not download-and-figure-it-out. An engineer wires it to your environment, validates end-to-end, and hands it off running.

One vendor, start to finish

Software, deployment, and ongoing support from the same source. No finger-pointing between vendors when something needs attention.

Perpetual licenses

You own what you buy. No annual subscription required to keep running software you've already paid for. Maintenance is optional, not a paywall.


Questions

Commonly asked.

Does ATLAS send our data to OpenAI, Google, or any AI cloud?

No. ATLAS runs a self-hosted language model on your own GPU hardware, inside your network. Your documents are indexed into a private knowledge base and never leave your walls. That is the entire point of an on-premise assistant.

What hardware does ATLAS need — and do you provide it?

ATLAS requires one or more dedicated enterprise GPUs (for example, NVIDIA A-series or L-series) with sufficient VRAM, plus a server to host them. We do not provide, sell, lease, or host the hardware, GPUs, virtualization, networking, or model licensing — that infrastructure is yours to supply and own. We're glad to help you spec an appropriate configuration before you buy.

How does ATLAS decide who can see what?

Every request carries the signed-in user's directory token. Your Active Directory / LDAP group membership determines which knowledge folders will answer — enforced on the server, with an audit trail. A person only gets answers from the documents their access already permits.

How is either product installed?

Entirely remotely. You provide secure remote access — VPN, RDP, SSH, or an equivalent — and our engineer installs, wires, and verifies the platform through it. No hardware shipments and no on-site visit are included. If your environment strictly requires on-site work, we can discuss that separately.

Why does ATLAS cost more than Server Guardian?

ATLAS is a fully on-premise AI platform: a self-hosted model, a retrieval pipeline over your documents, per-user access control tied to your directory, a mobile app, and an embeddable web widget — all deployed and tuned against your environment. Comparable enterprise on-prem AI/RAG builds run $80,000–$150,000 and up. Server Guardian is a focused monitoring front end and is priced accordingly.

Are these in the App Store or Google Play?

No. Both are distributed directly to your organization as compiled apps (and, for ATLAS, a web widget) after purchase. They stay off public listings, and you control distribution to your team.

Can we see them before buying?

Yes. We can share demo builds that run against simulated data — same code, no live infrastructure required — so you get a real feel for each product before any commitment. Reach out and we'll set one up.

Can they be branded for our organization?

Yes. Both can carry your organization's name, accent color, and logo — an internal tool that looks native to your environment, with no visible Bayou Byte Studio branding if that's what you need.

Let's talk about your environment.

Tell us what you're running — your monitoring stack, your document sources, your directory, the platforms you're targeting, and whether you already have GPU hardware for AI. We'll send back a demo build, a scoped quote, and a time to talk. No sales process — a direct conversation with the engineer who will do the work.

Ask about Server Guardian Ask about ATLAS

We read and respond to every message ourselves. · support@bayoubytestudio.com