Alteryx
Agent Stage 2 · Stage Verification pending

Account End of Year Planning Agent

Helps account teams plan the end of the year in Claude, using account, segmentation, and propensity data served through an MCP on DataHub and Snowflake.

Request a change
ABOUT THIS AGENT

What it does

Runs as an MCP over DataHub and Snowflake and is used in Claude through the plugins MCP. It draws on the segmentation work already shared with the TVPs, including vertical industry classification and propensity to buy, so planning starts from the same numbers as the operating model.

How it works

1A seller or planner asks for an account plan in Claude, in plain language.
2Claude calls the MCP through the plugins connector.
3The MCP reads account data from Snowflake and the definitions behind it from DataHub.
4It applies the FY26 account hierarchies, industry vertical, and propensity-to-buy scores.
5Claude drafts the end-of-year plan, and the planner reviews it before it is used.
AGENT
Account End of Year Planning Agent
TOOLS
DataHub
Snowflake
Plugins MCP
THE OUTPUT

What a plan looks like

A plan is read by the account team who act on it, by executives who allocate against it, and by analytics who check the model behind it. They all read the same plan. What makes that work is that every number carries where it came from and whether it has been signed off.

Every number names its source

A plan says which segmentation input each figure came from, so a reader can look it up in the semantic layer below.

Tied back to the operating model

The plan states where it matches the segmentation already shared with the TVPs, and where it has moved since.

Uncertainty is spelled out

Inputs still in validation are listed with what changes in the plan if they move on sign-off.

What a plan contains

Starting positionCurrent ARR, open pipeline, renewal date, and propensity to buy.
What it is built onEvery segmentation input used, its value for this account, and who owns it.
Tie-outWhere the plan matches the segmentation pack shared with the TVPs, and where it differs.
PrioritiesThe plays for the rest of the year, each naming the inputs it rests on.
What would changeWhich priorities move if an input still in validation changes on sign-off.
How it was builtDefinitions read from DataHub, model versions, what changed since the last run, and the analytics sign-off.
AN EXAMPLE

A worked plan, end to end

A full plan for one account, with sample values, showing how the parts above come together.

See an example plan

Agent details

WHAT IT KNOWS

Data and definitions

FY26 account hierarchies, vertical industry classification, and propensity-to-buy models. Each is defined in the semantic layer below.

EXAMPLE REQUEST

How to use it

“Build the end-of-year plan for [ACCOUNT].”

WHAT NEEDS A HUMAN

Approvals

The planner reviews every plan before it is shared.

WHO OWNS IT

Owner

[OWNER NAME]. Segmentation inputs are built by [ANALYTICS OWNER].

SUCCESS MEASURE

How we measure impact

Baseline: [BASELINE]. Target: [TARGET].

CONNECTIONS

What it connects to

DataHub · definitions and lineage
Snowflake · account and segmentation data
Claude plugins MCP · where planners use it
SEMANTIC LAYER

The definitions behind the agent

What each segmentation input means. Defined once, these keep planning, operating model design, and other agents on the same numbers.

Vertical industriesInitial industry classification is done and shared with the TVPs. Some field-driven refinement is expected, such as whether Professional Services stays within Tech and Tech Services or becomes its own vertical. AMER already uses it in org design.
Propensity to buyThe value and potential analysis, refreshed on the FY26 account hierarchies and shared with the TVPs. It is used in operating model design.
Disti vs directAnalysis of distributor-led versus direct accounts. An initial draft is targeted for validation the week of Oct 12.
WHO OWNS IT

Definitions and upkeep

[DEFINITIONS OWNER] owns the definitions. Maintenance and request intake sit with the GTM AI team.

WHERE IT GOES NEXT

Reuse across agents

The same segmentation definitions can serve other agents and Slack-based tools, so a vertical or a score means the same thing everywhere.

TO CERTIFY

Before verification

[RESOLVE: the Disti vs direct draft]

Where it is in the process

BEFORE

1 Propose

Scored on value, feasibility, and risk; data checked.

YOU ARE HERE

2 Stage

Built and tested on sample data; analytics and tech sign off.

NEXT

3 Pilot

Small group; a person approves every output.

THEN

4 Deploy

Wider rollout, with permissions and logging on.

Guardrails on every agent

Governed

Data comes through DataHub and Snowflake, not ad-hoc copies.

Reviewed

The planner approves every plan before it is shared.

Permissioned and logged

[CONFIRM: access controls and logging for the MCP]

Have a request for this agent?

Bring it to the GTM AI team. We'll point you to the right door.

Submit a request