Skip to content

Ignite AI Agents · CRM

Your data, answeredin plain language.

The Constituent Support agent answers questions about constituents, gifts, and KPIs straight from your live data. No report request, no query builder, no waiting. Every answer shows the sources behind it.

Read-only

It reads and cites your data. It never changes it.

Every answer carries

The answer first
The number, record, or list you asked for, up top
Date and filter context
The fiscal year and filters behind every total
Sources shown
The records or filters the answer was drawn from
Honest about gaps
“No results” and “I hit an error” are never blurred together

Ask your CRM anything.

Type a question the way you’d ask a colleague. The agent reads your live data and answers in plain language. It only ever reads.

Plain-language discovery

Ask by meaning, not by field. It searches notes and history for questions like “who has nursing experience” or “who has mentioned an interest in scholarships”, and shows the evidence for each match.

Who’s mentioned an interest in golf?

Financial answers with context

Totals, campaign giving, and gift counts, always with the date range and filters spelled out so the number means something.

What’s the total giving for the Summer 2025 campaign?

KPIs and trends, explained

Retention rate, new donors acquired, RFM scores, and giving trends, summarized in words instead of a raw export.

What’s our donor retention rate this year?

A full picture on one donor

Ask about a single constituent and get a briefing: giving, engagement, household, and the context that matters, pulled together.

Give me the full picture on Marguerite Osei.

Your saved queries, on command

Run the queries your team already built by asking for them in words. No rebuilding, no digging through menus.

Run our saved query for lapsed donors.

Answers you can open

Results come back as readable tables, lists, and KPI breakdowns, with direct links to open the record in Ignite when you want the source.

Show me our top 10 donors by lifetime giving.

How the agent works

Ask, find, answer, explore.

No query language to learn and no report queue to wait in. You ask, and the agent works your live data the way an analyst would.

  1. 01

    Ask

    Pose the question in plain language: a lookup, a total, a trend, or an open-ended who-has search.

  2. 02

    Find

    The agent picks the right read for the job: search by meaning, exact lookup, a single-donor briefing, or an analytics query.

  3. 03

    Answer

    The answer comes first, formatted for people, with the date range and its sources shown.

  4. 04

    Explore

    It offers the next cut: break this down by campaign, look at their giving history, widen the range.

Asked and answered, from live records.

Inside every answer

Answer first. Then the context.

The agent leads with what you asked for, then backs it up: the context, the filters, and the sources behind it. When the data is not there, it says so.

  1. 01Headline resultThe number, record, or list you asked for, first.
  2. 02Supporting detailThe detail behind the headline result.
  3. 03Filters and date rangeThe fiscal year and filters the answer used.
  4. 04Linked sourcesWhere the answer came from, with named records openable directly.
  5. 05When there is nothing to returnEmpty results and errors are named as such, never guessed past.

Want to see an answer built on your own data?

Book a working session

Governed by design

Read-only by default. Your admins set the rest.

It starts with read access and cites the records and filters behind every answer, which makes it the safest agent to put in front of your whole team. Admins control what it can read and do, and any write your team enables follows the permissions and approval gates you configure.

See the governance model

Read-only by default

Out of the box it answers questions and changes nothing. Where your admins enable a write, such as correcting a household relationship, it follows the approval gates your team configures. And by policy it will not bulk-dump records or export donor lists.

It does not make data up

Empty results are stated plainly. It never invents records or IDs, and never combines separate results to imply a relationship it did not verify.

Access follows your permissions

It reads within the permissions your organization already sets, and donor details are not carried between sessions.

It stays within the evidence

It distinguishes “no results” from “an error,” names the filters behind every total, and does not extrapolate beyond what the data shows.

No gates to widen here. This agent is read-only. It cites your data and never changes it.

Answers from the data you already keep.

No warehouse to build, no export to refresh. The agent reads the same live record your team works in every day, so the answer is current the moment you ask.

  • Gifts and pledges
  • Constituents and households
  • KPIs and analytics
  • Campaigns and funds

Available in the Intelligence Hub, embedded in constituent records, and in the chat panel inside Ignite.

Proven foundation

Built from the ground up on a modern, AI-first architecture.

Ignite is a new platform, built by the team nonprofits have trusted with their constituent and financial data for over 15 years.

G2 reviews

4.5/5

Average on G2.

20M+

Constituent profiles managed across the StratusLIVE platform family.

$1B+

In annual giving processed across the StratusLIVE platform family.

“One source of the truth about our contacts, donors, and partners that is easily kept up to date.”

Robert W.

Director of Fundraising Operations & Digital Solutions

Verified review · G2

“Removes the need for manual import and export of data.”

Charmaine V.

Small-business nonprofit

Verified review · G2

Honest questions

The questions teams actually ask about this agent.

Straight answers on trust, control, and your data. The full security posture, providers included, is on the Trust & Security page.

Can we trust an AI agent to answer from our data?

This is the safest agent to put in front of your whole team. It is read-only by default, so out of the box it answers and changes nothing. Where your admins enable a write, it follows the approval gates your team configures. It reads within the permissions your organization already sets, names the sources and filters behind every answer, and states an empty result plainly instead of papering over it.

Is this just ChatGPT with our data pasted in?

No. Nothing is pasted anywhere: the agent queries your live data inside Ignite and answers with the sources and filter context shown. A general chat tool would answer from general knowledge, could not cite your records, and could not tell finding nothing apart from hitting an error.

What happens when an answer is wrong?

The sources and filters behind each answer are shown at the point of reading, so a wrong answer is catchable before a decision, not after. And because it reads by default and writes only where your admins enable it, an imperfect answer changes nothing on its own.

Is our constituent data used to train AI models?

Prompts and outputs are not used to train foundation models by default. Donor details are not carried between sessions, and each task routes to a model matched to its data sensitivity. The Trust and Security page documents the full data flow.

Watch an answer come together.

Book a working session and see the Constituent Support agent read a live record, evidence and all.