Skip to main content
Admin42

Our specialists

Two specialists. One platform.

Admin42 builds a family of trusted AI specialists. Each one is built for a single role that exists in every IT department, does that job well, and calls on another specialist when it is the better fit. The human always stays in control. Today there are two.

NELDA · Network Engineer

What NELDA does.

NELDA — Network Engineer and Local Device Administrator

NELDA is a virtual network admin appliance for mission-critical networks. It plugs into the network, learns the environment, and helps with everything a network admin would do — in plain language, by conversation, on your premises.

NELDA works for the IT professional responsible for the network — not the end user. The everyday questions from staff are DAYNA's job. NELDA is the specialist behind the scenes.

Human-In-The-Loop by design. NELDA recommends. You approve. It executes. No autonomous action against your environment — ever.

nelda.local/admin

library

local backend

⌘K
We're getting reports of intermittent issues in the library.

local 0.4s · 58 tokens

Nothing in syslog stands out for the library segment right now. Can you tell me what people are seeing — dropped connections, slow speeds, a particular room or device? That will point me to the right place to look.
ask anything · drop a config to upload · ⌘↵ to send

Lab / development interactions

What NELDA is

A network administrator who happens to use AI.

NELDA is purpose-built for network administration. It runs on your premises, watches your gear, and — when you approve — executes the configuration changes it recommends. Narrow by design, accurate where it counts.

What NELDA isn't

A general-purpose AI or a search engine.

NELDA won't browse the web for you, write your essays, or answer trivia. Ask it the capital of the US and it'll tell you it doesn't know — and offer to route the question to a larger backend if you've configured one. That's the point.

Curious why we draw that line? Read about generative and agentic AI.

Honest uncertainty

Most AI always has an answer. NELDA knows when it doesn't.

Teaching an AI to recognize the edge of its own knowledge — and say "I don't know" instead of inventing something — is one of the hardest things to get right, and one of the most important. A model that always answers is also a model that will confidently tell you something wrong.

On a network, a confident wrong answer can take a building offline. NELDA would rather tell you the truth, say what it is sure of, and point you to the right tool for the rest. That restraint is a feature, not a limitation.

nelda.local/admin

new conversation

local backend

⌘K
What is the capital of the United States?

local 0.31s · 44 tokens

I don't know — that's outside my area. For a general question like that, a search engine or a general-purpose AI chatbot is the right tool. I'm focused on networking, system administration, and cybersecurity.
ask anything · drop a config to upload · ⌘↵ to send

Lab / development interactions

Capabilities

What NELDA does on day one.

  • 01 · Setup flow

    NELDA introduces itself.

    On first boot, NELDA reports its IP, gateway, and DNS — then walks you through pinning its address, configuring DNS & NTP, authentication (LDAP / AD or TACACS+ / RADIUS), and VM provisioning. Skip anything you're not ready for; NELDA will come back to it later.

  • 02 · Ask, integrate, or build

    NELDA fits the tools you already have.

    For every supporting service — NMS, DHCP, DNS, IPAM, configuration management, syslog, ticketing — NELDA first asks what you already run, integrates with it if you have one, and offers to build one if you don't. Your existing stack doesn't get replaced. It gets a teammate.

  • 03 · Discovery & monitoring

    NELDA finds your network.

    Passive discovery on power-up — ARP, DHCP leases, LLDP / CDP — with active scans (SNMP, ping sweeps, targeted probes) only after you approve them. NELDA maps devices, subnets, and existing tools, and surfaces what's actually worth your attention.

  • 04 · Recommend, approve, execute

    Human-In-The-Loop, every time.

    NELDA describes what it's seeing in plain language, proposes a fix, and shows the exact change it would make. You approve, defer, or skip. When you approve, it executes the change. When you defer, it waits. Nothing touches your environment without your go-ahead.

  • 05 · Honest uncertainty

    NELDA says "I don't know."

    NELDA is narrow by design — and built to detect its own uncertainty and surface it rather than guess. When it doesn't know, it tells you and offers to route the question to a bigger backend if one is configured. That same uncertainty signal is part of how NELDA decides whether to escalate — see how NELDA thinks below.

  • 06 · Security advisories

    NELDA watches for trouble that's actually yours.

    Continuous cross-referencing of CISA KEV, NVD / MITRE CVE, MITRE ATT&CK, and vendor PSIRTs against your discovered inventory. NELDA tells you which advisories apply to gear you actually run — patch path, mitigation, or workaround — never generic alerts.

  • 07 · Scales with you

    One interface, more horsepower.

    NELDA runs locally out of the box. When you need more capacity or specialized skills, expand on your terms — another GPU, a dedicated GPU server, an all-in-one box like an NVIDIA DGX Spark or AMD Strix Halo, or an external LLM. We favor the all-in-one path for its simplicity, but the choice is yours.

  • 08 · Curated, fine-tuned models

    The right model for the job.

    We've selected the best models for each kind of task and fine-tuned them for network administration. NELDA also uses retrieval-augmented generation over publicly available documentation from the major networking vendors. When you have authorized access to official vendor documentation, NELDA can be further trained on your materials for environment-specific accuracy.

  • 09 · On-premise & private

    Your data stays yours.

    NELDA runs on your network. Telemetry, configurations, and logs do not leave your environment unless you explicitly choose to route a question to an external backend. Privacy is structural, not promised.

How NELDA thinks

Four levels. One conversation. The human is always last.

NELDA escalates only as needed. Routine work stays on the appliance; harder questions reach for bigger models; every path that touches your environment ends with you.

  1. Level 1 · Always present

    Onboard LLM

    A small, fine-tuned model running on the NELDA appliance itself. Handles routine Q&A, basic configuration suggestions, and most day-to-day work locally.

  2. Level 2 · Optional

    Specialized LLM on your hardware

    Larger or vendor-specialized models running on your own hardware — DGX Spark, Strix Halo, a GPU server you build. NELDA orchestrates loading, routing, and use.

  3. Level 3 · Optional

    External LLM

    Cloud-hosted models reached over the API — Gemini, ChatGPT, Claude — when extra capacity or capability is worth the round-trip. Off by default; on only if you turn it on.

  4. Level 4 · Always

    The human

    You are the final authority on anything that touches your environment. Every recommendation from L1, L2, or L3 ends here for your approval. That's Human-In-The-Loop.

Common escalation paths

NELDA picks the path that fits the task.

  • Routine: L1 → L4 (onboard answers, you approve)
  • Specialized: L1 → L2 → L4 (onboard hands off to your bigger model, you approve)
  • External help: L1 → L3 → L4 (onboard reaches an external LLM, you approve)
  • Full chain: L1 → L2 → L3 → L4 (when the question warrants every tool)

Human-In-The-Loop, in action

NELDA recommends. You approve. It executes.

nelda.local/admin

guest wifi

local backend

⌘K
Is the guest WiFi setup alright?

local 0.5s · 62 tokens

The guest VLAN has no rate limit, so it can crowd out staff traffic at busy times. I'd add a 20 Mbps cap on VLAN 30 — here's the exact change. Approve it and I will apply it.
policy-map GUEST-LIMIT
class class-default
+ police cir 20000000 bc 625000
interface Vlan30
+ service-policy input GUEST-LIMIT
Yes, go ahead.

local applied · just now

Done. The 20 Mbps cap is live on VLAN 30, and I've logged the change. I will keep an eye on staff throughput and let you know if anything looks off.
ask anything · drop a config to upload · ⌘↵ to send

Lab / development interactions

DAYNA · Smart Desktop Services

Meet DAYNA, first-line help for everyone else.

DAYNA — Digital Assistant for Your Networked Applications

Where NELDA helps the people who run the network, DAYNA helps the people who use it. It answers everyday questions, solves common desktop problems, gathers diagnostics, and guides people through routine issues — often before they ever become a help desk ticket.

  • Password help
  • Printer problems
  • VPN connectivity
  • Wi-Fi troubleshooting
  • Application access
  • Basic software support
  • "How do I…" questions
dayna.local/help

DAYNA

smart desktop services

⌘K
I'm locked out of my email.

dayna just now

I can help with that. Your password expired yesterday, which is why you're locked out. I can text a reset link to the phone on file, or walk you through resetting it here. Which would you prefer?
Text me the link, please.

dayna sent · just now

Sent. It's good for 15 minutes. Once you've set a new password, give it a minute to sync and your email will sign in again. Anything else I can help with?
ask anything · attach a file or screenshot · ↵ to send

Lab / development interactions

When it's really the network

DAYNA knows when to bring in NELDA.

Some problems look like a desktop issue but are really an infrastructure issue. When DAYNA sees one, it gathers the relevant details and consults NELDA. NELDA analyzes the network, returns its findings for a human to approve, and DAYNA explains the result back in plain language.

DAYNA is a different specialist built on the same platform as NELDA. They share the orchestration underneath — which is how they hand work back and forth without either one stepping outside its own role.

dayna.local/help

DAYNA

session 20260628-180700

⌘K
The internet is slow.

dayna just now

Thanks for flagging it. It looks like you're connecting over Wi-Fi from the Middle School — is that right?
Yes — and a few of us. Video calls keep dropping.

DAYNA checks with NELDA

DAYNA shares the details with NELDA. NELDA finds the Middle School uplink saturated by a backup job running in the middle of the day, and logs a recommendation for the Technology Administrator to move the backup to overnight.

Dramatization of the AI-to-AI exchange

dayna just now

Here's what's happening: a large backup is using up the Middle School's connection during the day, which is what keeps interrupting the video calls. I've logged it with NELDA, and NELDA will raise it with the Technology Administrator along with a recommendation to move the backup to overnight.

dayna just now

Your reference is Chat Session 20260628-180700 if you need to follow up. Would you like me to email you a summary?
ask anything · attach a file or screenshot · ↵ to send

Lab / development interactions

A growing family

One job each. Working together when it counts.

Admin42 is building a family of trusted AI specialists, each with one clearly defined job. NELDA and DAYNA are the first two. More will follow for security, systems administration, backup, monitoring, and the other roles that keep an organization running. Every specialist owns its area, knows when another is better suited, and leaves the operational decisions to the human in charge.

  • NELDA

    Network engineer — available now

  • DAYNA

    Smart desktop services — available now

  • Security

    In the family roadmap

  • Systems administration

    In the family roadmap

  • Backup

    In the family roadmap

  • Monitoring

    In the family roadmap

Want to see the specialists in your environment?

We work with a small number of early customers at a time. If your network sounds like the kind of place NELDA was built for, we'd like to hear about it.