Enhanced AI Addon

Enhanced AI

Name: Resgrid Enhanced AI
Brand: Resgrid
Price: 95 USD

Private AI that answers from your own Resgrid data. Admin Assist AI with no question cap, report narrative drafts, call briefs, catch-up summaries and cited answers from your documents, running on open-weight models on Resgrid-owned GPUs and never sent to cloud AI providers. Or bring your own AI provider. People stay in charge of every record.

Admin Assist AI

Example

> Why wasn't J. Sample paged
  for call #1042?

  Call #1042 was dispatched to
  the Station 2 group. J. Sample
  is only in Station 1, so no
  page went to them.

  Sources  Call #1042 dispatch,
           Station 2 members
  Open     Groups  ->

  No settings were changed.

Illustrative example with made-up names and data.

Open-Weight Models. Our Hardware or Yours. Never Cloud AI.

Resgrid will never send your department's data to a cloud LLM provider.

Open-weight models only

Enhanced AI runs openly published models, currently Qwen3-14B with Qwen3-8B as a fallback, not a closed model behind another company's API.

On hardware we own, or on yours

The models run on GPU servers Resgrid owns and operates in your Resgrid region. Self-hosting Resgrid? Point its AI at open-weight models on your own private hardware.

Never sent to cloud AI providers

Resgrid never sends your department's data to cloud AI providers such as OpenAI, Anthropic, Google or Microsoft Azure, and never uses it to train models.

The only way your data reaches an outside AI provider is if your department connects its own provider account. That is your choice, and it is off by default.

What Every Department Already Gets

Resgrid's everyday AI and automation tools are free on every plan. Enhanced AI is not a paywall around them. It is the model-backed layer on top: answers and drafts grounded in your own data, for the whole department.

Chatbot built-in commandsIncluded
Incident board narratorIncluded
ICS playbooksIncluded
Run cardsIncluded
Keyword searchIncluded
Setup WizardIncluded
Setup ReportIncluded
Admin Assist health checksIncluded
MCP serverIncluded
Text-to-speechIncluded
On-device voice-to-textIncluded

Free Admin Assist AI questions

Included

Every department can try Admin Assist AI without buying anything: 20 questions during the first 30 days after your department's first answered question, then 4 questions every calendar month. Unused questions do not roll over, and a question only counts when Admin Assist gives an answer. Enhanced AI removes the question cap, and answers draw from your department's monthly AI budget instead.

Commands

Included with every plan

Chatbot built-in commands, run cards, ICS playbooks and the incident board narrator

Covers: Getting routine work done by message or on the board

Find and set up

Included with every plan

Keyword search, Setup Wizard, Setup Report and Admin Assist configuration health checks

Covers: Finding things and spotting configuration problems

Voice and connect

Included with every plan

Text-to-speech, on-device voice-to-text and the Resgrid MCP server

Covers: Hands-free use and connecting compatible AI assistants

Try Admin Assist AI

Included with every plan

20 Admin Assist AI questions in your first 30 days, then 4 every calendar month

Covers: Asking an administrator's question in plain language

Enhanced AI

Enhanced AI addon

Admin Assist AI with no question cap, assistant and incident questions, report drafts, call briefs, catch-up summaries, knowledge and analytics answers

Covers: Grounded answers and drafts, every day, for the whole department

Adding Enhanced AI does not change anything above it. The free tools keep working exactly as they do today, with or without the addon.

What Enhanced AI Adds

Seven capabilities available at launch, each grounded in your department's own Resgrid data and built to hand the final word to a person.

Available at launch

Admin Assist AI

A conversational helper only for department administrators, inside the Admin Assist page. It answers from your department's own configuration and Resgrid's documentation, and works alongside the free Setup Wizard, Setup Report and configuration health checks.

Explains any setting in plain language
Works out why something isn't working, such as why a member wasn't paged
Tells you what a change will affect before you make it
Walks you through setting Resgrid up for volunteer or career fire, EMS, search and rescue, emergency management and more
Answers from your department's own configuration and Resgrid's documentation
Links straight to the right screen and never changes settings itself

Available at launch

Assistant and Incident Questions

Ask the Resgrid assistant free-form questions, and let incident commanders ask about the live command board: "Who is overdue for PAR?" "What is unassigned?"

Available at launch

Report Narrative Drafts

Incident report narrative drafts for NERIS and NFIRS from the call, timeline, units and notes, plus after-action report prose and ICS-201 or transfer-of-command briefings. Drafts only; your review and signature steps stay.

Available at launch

Call Briefs

One readable brief for responders, built from the pre-plan, hazards, shutoffs, hydrants, weather alerts, prior calls at the address and the run-card assignment.

Available at launch

Catch-Up Summaries

"Catch me up" on a chat thread or on the last stretch of an incident timeline, instead of scrolling back through everything that happened while you were away.

Available at launch

Knowledge Answers

Ask about your own documents, protocols and notes, such as "What is our mutual aid policy?", and get an answer with citations back to the source.

Available at launch

Analytics Questions

Ask plain-language questions about response times and workload. Answers come from Resgrid's reporting data, never from free-form database queries.

Coming Later

Planned for Enhanced AI after launch. These are not available yet, and we are not announcing dates.

Coming soon

AI Dispatch Assistance

Help reading inbound CAD emails and texts, with automatic fallback to standard processing.

Coming soon

Dispatch Audio Transcription

Transcripts of dispatch audio alongside the call.

Coming soon

Semantic Search

Search by meaning across Resgrid modules, not just by keyword.

Coming soon

Push-To-Talk Transcription

Text transcripts of push-to-talk transmissions.

Admin Assist AI: Answers for the People Who Configure Resgrid

Admin Assist AI is a conversational helper for department administrators, inside the Admin Assist page. Ask why something is behaving the way it is, what a change would affect, or how a department like yours usually sets Resgrid up. It answers from your department's own configuration and Resgrid's documentation, shows where the answer came from, and links you straight to the screen where you can make the change yourself.

Available only to department administrators
Explains settings, troubleshoots and previews the impact of a change
Guidance for fire, EMS, search and rescue, emergency management and more
Links to the right screen instead of changing anything for you
Works alongside the free Setup Wizard, Setup Report and health checks
Free questions every month for every department; no cap with Enhanced AI

Drafts That Start From the Record, Not a Blank Page

The call, the timeline, the units and the notes are already in Resgrid. Enhanced AI uses them to draft the incident report narrative, the after-action prose or the ICS-201 and transfer-of-command briefing, so the person writing starts by reviewing and editing instead of typing from memory. Nothing becomes a record until a person has reviewed, edited and signed it.

Incident report narrative drafts for NERIS and NFIRS
After-action report prose
ICS-201 and transfer-of-command briefings
Built from your call, timeline, units and notes
Your existing review and signature steps stay in place
A person signs anything that becomes a record

One Brief Instead of Six Screens

Before arrival, the useful information about an address is spread across the pre-plan, hazards, shutoffs, hydrants, weather alerts, previous calls and the run card. A call brief pulls it together into one readable summary for responders. Catch-up summaries do the same for a busy chat thread or the last stretch of an incident timeline.

Pre-plan, hazards, shutoffs and hydrants in one place
Weather alerts and prior calls at the address
The run-card assignment for the call
"Catch me up" on a chat thread or incident timeline
Grounded in your department's own Resgrid data
Shows where the information came from

People Stay in Charge

AI answers questions and writes drafts. A person reviews, edits and signs anything that becomes a record. AI never changes department configuration, and any AI step in an emergency workflow has a short deadline and a standard fallback.

Someone asks

An administrator, incident commander or member asks a question or requests a draft.

Your people

It reads your data

Enhanced AI works from your department's own Resgrid data and Resgrid's documentation.

Enhanced AI

It answers or drafts

Every answer shows where it came from. Admin Assist AI links to the screen that matters.

Enhanced AI

A person decides

Your people review, edit and sign anything that becomes a record, and make any configuration change themselves.

Your people

Your process stays

Validation, signatures and quality review are unchanged. AI drafts never become records until a person accepts them.

Your people

Private by Design

Like the rest of Resgrid's AI, Enhanced AI keeps your department's data inside Resgrid. Here is exactly what that means.

Resgrid-Owned GPU Servers

Enhanced AI runs open-weight models on GPU servers that Resgrid owns and operates. You do not need your own GPU, AI account or API key.

Never Sent to Cloud AI

Resgrid never sends department data to cloud AI providers, and there is no dependency on OpenAI, Anthropic, Google or Azure. Want one anyway? Bring your own provider.

Stays in Your Region

AI processing stays in your Resgrid region, so a department on EU Central is processed in EU Central.

Not Used for Training

Enhanced AI does not use your department's data to train AI models.

Want the detail? See how Resgrid secures AI and your data.

The Models We Run

Every model is open-weight, runs on hardware Resgrid owns and operates, and is reviewed and pinned to a specific version before your department's data reaches it.

Model	What it does	Status
Qwen3-14B	Admin Assist AI, assistant answers, report and briefing drafts, call briefs	Available at launch
Qwen3-8B	A smaller fallback for the same work	Available at launch
BGE-M3	Embeddings for semantic search	Coming soon
Whisper large-v3-turbo	Dispatch audio transcription, run with faster-whisper	Coming soon

Included with Enhanced AI

Bring Your Own AI Provider

Already pay for an AI provider? Connect your department's own subscription and the Resgrid assistant uses it instead of Resgrid's models. Every popular provider works.

OpenAI
Anthropic (Claude)
Google Gemini
Microsoft Azure OpenAI
Mistral AI
xAI (Grok)
DeepSeek
Groq
Together AI
Fireworks AI
OpenRouter
Perplexity
Cohere
Any OpenAI-compatible endpoint

Your subscription, your agreement

Connect your department's own provider account with an API key. The provider bills you for that usage under your agreement with them; it does not draw from your Enhanced AI budget.

Off until you turn it on

Nothing goes to an outside provider until a department administrator connects one. Remove it at any time and the assistant goes back to Resgrid's own models. It is never available with Advanced Data Protection.

Powers the Resgrid assistant

Your provider handles the Resgrid assistant's language understanding and conversational replies. Admin Assist AI, drafts and call briefs stay on Resgrid's open-weight models.

Key encrypted, endpoint checked

Your API key is encrypted at rest and never shown again after you save it. Provider endpoints must use HTTPS and a public address.

Your own provider is the one exception to Resgrid never sending data to cloud AI, and it is your department's decision. Messages sent to it are handled under your agreement with that provider. Departments using Advanced Data Protection cannot connect one, so their protected data never leaves Resgrid.

Compared to Pasting Data Into a General AI Chat Tool

A general-purpose chat tool only knows what someone copies into it. Enhanced AI works from the data your department already keeps in Resgrid, and keeps that data where it is.

Topic	A general AI chat tool	Resgrid Enhanced AI
Where your data goes	Copied out of Resgrid into a separate AI service	Processed by open-weight models on Resgrid-owned GPU servers in your Resgrid region, never sent to cloud AI providers
What it knows	Only what someone pastes into the chat	Your department's own Resgrid data and Resgrid's documentation
Showing its work	No link back to the record the answer came from	Every answer shows where it came from
Getting to the fix	You still have to find the right screen yourself	Admin Assist AI links straight to the right screen
Records	Output is copied and pasted into your report	Drafts built from the call, reviewed and signed by your people
Paying for it	Often priced per user or by usage	One price per department, with a monthly AI budget and no overage charges

Enhanced AI Pricing

Enhanced AI is a per-department addon billed monthly. It is priced per department, not per user, and it requires a paid Resgrid plan. Only your department's managing member can buy or cancel it.

US-West (Worldwide)

$95 /mo

per department, billed monthly

EU Central (EU)

€145 /mo

per department, billed monthly

A monthly AI budget

Each department gets a monthly AI usage budget. If you reach it, AI features fall back to the standard non-AI experience until the next month.

No overage charges

Reaching the budget never adds to your bill. No overage charges, no surprise bills.

Cancel any time

Cancel whenever you like. Access continues to the end of the month you have paid for.

Every department gets free Admin Assist AI questions without buying Enhanced AI. The annual-plan discount applies to your subscription, not to addons.

What Enhanced AI Does Not Do

AI in emergency services is only useful if you know exactly where it stops. These are the deliberate boundaries.

No autonomous dispatch decisions

At launch, Enhanced AI answers questions and writes drafts; it does not dispatch. The dispatch assistance coming later works inside your existing dispatch rules, with a short deadline and automatic fallback to standard processing, so a slow or unavailable AI never stops a call from being created.

No configuration changes

Admin Assist AI explains settings and links you to the right screen. Any change to your department's configuration is made by a person.

No free-form database queries

Analytics questions are answered from Resgrid's reporting data, not by running open-ended queries against your database.

It does not replace your review

Drafts are drafts. Your validation, signatures and quality review stay exactly where they are, and a person signs anything that becomes a record.

Not a medical or clinical decision tool

Enhanced AI does not make or recommend clinical decisions. Patient care decisions stay with your clinicians and your protocols.

Not a compliance certification

Buying Enhanced AI does not certify your department against any standard. Report drafts still need to meet your reporting requirements when reviewed.

Looking for the free chatbot and natural language commands? See AI & Natural Language, included with every plan.

Enhanced AI Questions

What does Enhanced AI include?

Admin Assist AI for department administrators with no question cap, free-form questions to the Resgrid assistant and the live incident command board, incident report narrative drafts (NERIS/NFIRS), after-action prose and ICS-201 or transfer-of-command briefings, call briefs, catch-up summaries, knowledge answers from your own documents with citations, and plain-language analytics questions. It includes a monthly AI usage budget for your department.

Is our data sent to OpenAI or other AI companies?

No. Resgrid never sends your data to cloud AI providers. Enhanced AI runs open-weight models on GPU servers Resgrid owns and operates, and AI processing stays in your Resgrid region. There is no dependency on OpenAI, Anthropic, Google or Azure, and your data is not used to train AI models. The one exception is your own choice: if a department administrator connects the department's own AI provider subscription, the Resgrid assistant sends that department's messages to the provider they chose.

Which AI models does Resgrid use?

Open-weight models only. Today that is Qwen3-14B, with Qwen3-8B as a smaller fallback, for Admin Assist AI, answers, drafts and call briefs. Semantic search will use BGE-M3 embeddings and dispatch audio transcription will use Whisper large-v3-turbo when they arrive. Resgrid reviews and pins each model version before it is used.

Can we use our own AI provider subscription?

Yes, with Enhanced AI. A department administrator can connect the department's own account with OpenAI, Anthropic (Claude), Google Gemini, Microsoft Azure OpenAI, Mistral AI, xAI (Grok), DeepSeek, Groq, Together AI, Fireworks AI, OpenRouter, Perplexity, Cohere or any OpenAI-compatible endpoint, including a model server you run yourself. The Resgrid assistant then uses that provider for the department's messages, under your agreement with the provider, who bills you for that usage. It is off until you turn it on, and it is not available to departments using Advanced Data Protection, so their protected data never leaves Resgrid.

Can we run the models on our own hardware?

Yes, if you self-host. Resgrid is open source, and a self-hosted installation can point its AI at open-weight models on your own private hardware. On Resgrid's hosted service the models run on hardware Resgrid owns and operates, and with Enhanced AI you can also point the Resgrid assistant at a model server you run, as long as it is reachable over HTTPS.

What happens when we reach the monthly AI budget?

AI features fall back to the standard non-AI experience until the next month. There are no overage charges and no surprise bills.

Can we try Admin Assist AI for free?

Yes. Every department gets 20 Admin Assist AI questions during the first 30 days after its first answered question, then 4 questions every calendar month. Unused questions do not roll over, and a question only counts when Admin Assist gives an answer. Enhanced AI removes the question cap; answers then draw from your department's monthly AI budget.

Do we need our own GPU or AI account?

No. Resgrid operates the GPU servers and the models. There is nothing to install and no separate AI account or API key to set up. If you would rather use an AI provider you already pay for, Enhanced AI lets you connect it.

Who can buy it?

The department's managing member can buy or cancel Enhanced AI, and it requires a paid Resgrid plan. It is priced per department, not per user, billed monthly, and you can cancel at any time; access continues to the end of the paid month.

Grounded in the Data You Already Keep

Enhanced AI answers from your department's own Resgrid data, including configuration, calls, timelines, pre-plans, documents and notes, and shows where each answer came from.

Admin Assist

Resgrid Assistant

Incident Command

Calls & Timelines

Pre-Plans & Hazards

Run Cards

Chat

Documents & Notes

Reporting

Put Your Own Data to Work, Privately

Add Enhanced AI to your paid plan and give your department grounded answers and first drafts, on Resgrid-operated servers, with people signing off on every record.

Per Dept

Not Per User

Zero

Overage Charges

Your Region

Where AI Processing Stays

4 / mo

Free Admin Assist Questions