Create the workspace
A human owner starts the free trial, chooses GTM Engineer during onboarding, and connects the systems AmpUp should understand.
AmpUp for GTM Engineers
Give your agent a GTM operating system.
You create the workspace. Your agent connects to AmpUp and gets governed access to the meetings, transcripts, deal intelligence, scores, tasks, and CRM context it needs to do useful work.
14-day free trialNo credit cardOAuth or API key
agent.ts
// Pull analyzed calls for the agent
const meetings = await ampup(
"/v1/list_meetings",
{ status: "analyzed", limit: 25 }
);
// Then retrieve evidence, not guesses
const meeting = await ampup(
"/v1/get_meeting",
{ meeting_id: selectedMeetingId }
);
Sales stackAmpUp contextYour agent
A clean handoff
You sign up. Your agent goes to work.
The human owns the workspace and consent. The agent gets a clear, permissioned interface to the operating context.
Authorize the agent
Use OAuth for an MCP client or create a revocable API key for a server-side integration. Every request stays scoped to your workspace.
Run the workflow
Let the agent retrieve context, assemble evidence, and draft the next action. Keep material writes behind review and approval.
Two interfaces, one context layer
Connect naturally. Automate deterministically.
MCP is ideal for interactive agents that discover and compose tools. The direct API is ideal for scheduled jobs, ETL, and services with a fixed contract.
See requests, responses, and production patternsAgent-native
Connect over MCP
Add the hosted AmpUp server to Claude, Codex, Cursor, or any MCP-over-HTTP client. Sign in with OAuth; no API key is needed for compatible clients.
claude mcp add --transport http ampup https://app.ampup.ai/mcp
MCP setup and tool referenceService-native
Call the API directly
Create a key under Settings → API Keys, keep it server-side, and POST JSON to the AmpUp API. Use this path for reliable, repeatable data movement.
curl https://app.ampup.ai/mcp/api/v1/list_meetings \
-H "Authorization: Bearer $AMPUP_API_KEY" \
-H "Content-Type: application/json" \
-d '{"status":"analyzed","limit":25}'
The context your workflows need
Pull evidence from the whole revenue loop.
AmpUp turns scattered GTM systems into records an agent can retrieve, reason over, and use to prepare the next action.
Meetings & transcripts
Search analyzed meetings, retrieve summaries, and pull exact transcript excerpts only when the workflow needs evidence.
speaker turnschapterskey moments
Accounts, contacts & deals
Resolve the customer record, stakeholder map, pipeline stage, open tasks, notes, and connected CRM context.
deal stateownersrelationships
Scores & deal intelligence
Pull configured evaluation metrics, coaching feedback, objections, qualification evidence, and deal-level trends.
custom metricsfeedbackMEDDPICC
Prospecting & coaching
Work with audiences, campaigns, sequences, roleplays, assignments, and sessions through the governed tool surface.
ICP fitdraftspractice
Five journeys, ready to build
From signal to next action.
Start with reads and drafts. Add approvals at the boundary where the workflow changes a system or contacts a person.
01
The CRM never goes stale
A meeting finishes
Pull the analysis → resolve account, contact, and deal → draft field updates and follow-up tasks → ask for approval.
CRM changes backed by what the customer actually said.
02
List building at GTM speed
A segment or territory opens
Search companies and people → enrich missing data → score ICP fit → save an audience → draft a sequence.
A reviewable prospecting package without another spreadsheet handoff.
03
Conversation data reaches BI
The warehouse job runs
List newly analyzed meetings → bulk-pull scores and deal intelligence → normalize the payload → load the destination.
Meeting evidence joins pipeline data in the metrics layer.
04
Coaching closes the loop
A skill score drops
Find the weak call moment → build a meeting-derived roleplay → propose an assignment → read back practice results.
The next practice session is tied to a live deal, not a generic script.
05
Signals become timely outreach
A product, hiring, or intent signal arrives
Reconcile the company and people → add the signal as context → draft signal-aware outreach → queue human review.
Relevant outreach while the signal is still fresh.
Guardrails built into the handoff
Read broadly. Act deliberately.
Your identity, your workspace
OAuth and API credentials are resolved to the authorizing user and organization on every request.
Revocable access
Use short-lived OAuth for interactive agents or a separately managed API key for server-side jobs.
Human approval for side effects
Treat CRM sync, campaign launch, and outbound messaging as review gates, not background defaults.
Frequently asked questions
Build the first workflow with confidence.
Can an AI agent create its own AmpUp account?
A human owner starts the workspace and authorizes access. After that handoff, the agent can connect through OAuth over MCP or use a server-side API key. This keeps ownership, billing, and consent attributable to a person or team.
Should I use MCP or call the API directly?
Use MCP when an interactive agent should discover tools and choose the right operation from natural language. Use direct API calls for deterministic jobs such as warehouse loads, scheduled reports, or a service you already operate. Both paths use the same AmpUp data and permissions.
What is the best first read for an agent?
Start with the workspace overview, then list the relevant records. For meetings, retrieve the meeting overview before requesting a full transcript; the overview already includes analysis and tells the agent which deeper artifacts are available.
How should writes be handled?
Separate evidence collection from side effects. Let the agent read and draft freely inside its permissions, then require a clear approval before CRM sync, campaign launch, sending, or any action that affects a customer-facing system.
Give your agent better context than a browser tab.
Start a workspace, connect your agent, and build the first GTM workflow against real customer evidence.