GTM strategy & operations · annual planning

Dylan Ram

Systems at scale, across the whole revenue function.

Built out the foundation of Databricks' Partner Strategy & Ops team: forecasting, attribution, incentives, quotas, planning — and AI deployed on top of all of it.

// the work, two ways

The operating spine of a GTM org — zero to one, then one to 100. I run the forecast, attribution, incentive, and quota systems a partner business depends on, and align them across a two-sided marketplace.

This page runs its own eval suite: every answer it gives is pinned by tests. Repo →

Signature work

Systems that didn't exist before — built zero to one, scaled one to 100, and run across a two-sided marketplace: the partner team, the sellers who co-sell through partners, and the partners themselves.

The forecast

Gave the partner business a number it could plan against. Still the number the org runs on today.

The attribution model

Settled how partner revenue gets counted: one standard the whole marketplace aligns to, still in force.

The new-logo incentive

Turned partner incentives into a growth lever for new business. Still running.

The rules of engagement

Unblocked the partner channel as a revenue source: lead flow that moves in days, not weeks.

Territory designZero-to-one builds Strategic initiativesAnnual & headcount planning Executive partnershipLeadership presentations

The agentic system

Stood up at Databricks partner ops and still running: reporting, partner recommendations, and natural-language answers, from raw GTM data to agents reps actually use.

Sources
Salesforce · Spark & SQL pipelines
Governed data
Medallion data layer · one metric dictionary
Agents
Reporting agents · partner recommendations for reps · partner-fit Q&A · self-serve analytics
The org
Partner team · sellers · leadership, answering their own questions

The operating rule, everywhere: the money path never touches a model. Plain-English rules → deterministic, tested code → evals pinning the math.

Built in the open

This repo demonstrates the same discipline, built from scratch on synthetic data: five layers, from governed data to a loop that runs the QBR on the tooling itself.

Memory
what every session boots knowing
Data
one governed dictionary, defined once
Big rocks
a living plan per initiative
Skills
recurring work, made invocable
Loop
the setup reviews and rewrites itself

This isn't a toy domain: incentive crediting is a class of problem I know from running it — and everything here is built from scratch on synthetic data. The rule that makes AI safe in production: the model authors rules in plain English; deterministic, tested code computes anything that touches money. An eval suite pins the math — mid-quarter hires, territory handoffs, split credit, coverage gaps — so no model ever computes a credited dollar, and no deal is silently zeroed.

Experience

2021 — now
Partner Strategy & Ops Manager · Databricks
first Partner S&O hire · promoted 2023
Built the data-and-AI foundation and the forecast, attribution, and incentive systems the partner business runs on; deployed the team's first LLM agents; leads quota-setting across Sales, Finance, and Partner leadership.
2019 — 2021
SMB Sales Strategy & Operations Analyst · Salesforce
Built the territory-carving model behind the annual GTM plan, automated QBR and forecast-accuracy tooling, and was the direct analytics partner to a $250M AMER SMB business.
2018 — 2019
Business Data Analyst · CBRE
Owned the product-analytics stack end to end — data warehouse, Python data collection, and client-facing Tableau dashboards.

Take the résumé

One résumé, the whole record. The ATS-safe version is single-column for application portals; the default is designed for reading.

The stories behind these systems are better told live.