facebook-pixel
Field ops on autopilot →AI Dispatcher → 1:1 demo

Research Agent

The Research Agent is an AI agent in FieldCamp you invoke on demand with any operational or business question — it reasons over your FieldCamp data, searches the web when the answer lives outside, and returns a sourced, structured answer instead of a guess.

Runs on demandReads your dataSearches the web

A real template from the FieldCamp marketplace, configured on your operation. Setup included in your plan.

About this agent

Where it came from

Ships in the FieldCamp agent marketplace — the general-purpose analyst for every question the other agents weren't built for.

Every week produces questions that die unanswered because answering means an hour someone doesn't have: which ZIP codes drive our best margins? What's a fair price for hydro-jetting in this market? Did the spring campaign customers stick around? The data exists — in your records, on the web — behind an hour of assembly.

The Research Agent is the hour, on demand. Ask in plain language; it decides what to pull from your FieldCamp records, what to search outside, and how to put them together — and answers with its sources shown. The questions you stopped asking become questions you just ask.

What it actually does

Trigger: Manual — ask it anything

  1. 1

    Takes your questionreceive_question

    Plain language, asked in chat — operational, financial, or market-facing.

  2. 2

    Mines your recordsquery_fieldcamp_data

    Jobs, clients, invoices, visits — whatever the question touches internally.

  3. 3

    Searches the outsideweb_search

    Market rates, benchmarks, suppliers — when the answer isn't in your data.

  4. 4

    Answers with sourcescompose_answer

    A structured answer that shows where every claim came from.

What you get

Sourced answers to the questions that used to die on the someday list — in minutes, on demand.

A run, as you’d see it

Agent runs land on a timeline — what fired, what the agent found, and the action waiting for a human. This is that screen.

Question asked

Manual · via chat

90% confidencePending

Details

Question

"Best-margin ZIP codes this year?"

Summary

Top margin ZIPs: 89117 (38%), 89134 (35%), 89052 (33%) — driven by maintenance-plan density, not job size. Worst: 89030 at 14%, drive time eats it. Full table with job counts and sources attached.

DismissOpen answer

Answers questions; takes no actions. The safest agent in the library to hand the whole team.

By trade

Same agent, configured to how your vertical actually works.

HVAC

Equipment-spec lookups mid-job: "what's the refrigerant for this model?" — answered at the curb.

Landscaping

Plant and material sourcing questions hit the web with your job context attached.

Questions, answered

What kinds of questions can it handle?

Anything answerable from your records plus the public web: operational ("average days from estimate to acceptance?"), financial ("margin by job type last quarter?"), market ("going rate for drain camera inspections here?"). It tells you when a question needs data neither source has.

How do I trust the answers?

Every answer shows its work — which records it queried, which sources it read, links included. Treat it like a sharp analyst: check the sources on the answers that drive big decisions.

Can my whole team use it?

Yes — it respects FieldCamp's role permissions, so each person's questions run against what they're allowed to see. Techs ask spec questions; owners ask margin questions.

Have an agent idea we haven’t built?

The ideas section of this library exists because customers keep asking "could it just…?" Bring yours. If it should exist, we build it — and it ships as a template like this one.

No demo deck. Just your business and ours.