For GTM engineers
Wire GTM logic into your stack, not your CRM
You build the systems that run GTM. Gradient Works puts territory and routing logic behind an API your agent can call directly.
01The shift
Before and after Gradient Works
Before Gradient Works
- Territory logic lives in spreadsheets no API can reach.
- Routing rules break and nobody can explain why.
- Every new tool needs its own custom integration.
- Stakeholders want more scenarios than you have time to build.
- Your agent can read Salesforce data but can’t act on it.
After Gradient Works
- Territory and routing logic live behind one API.
- Your agent explains any routing decision in seconds.
- MCP connects your AI tools with no new integration work.
- Agents generate and compare scenarios on demand.
- Your agent can act on Gradient Works data, not just read it.
02What you get
What GTM engineers get with Gradient Works
- An interface, not a UI
Territory design, routing configuration, flow management and rep metrics are all addressable. The MCP exposes over 50 skills and tools, so the platform is scriptable rather than clickable.
- Debugging that takes seconds
Ask why a specific record was assigned the way it was and get the matching logic, the rule that fired and the assignment history in one answer. This is the capability GTM engineers notice first.
- No new integration surface
Any MCP-speaking agent connects — Claude Desktop, Claude Code, ChatGPT, OpenAI Codex, Cursor — without you building and maintaining another connector.
Most GTM tooling treats an API as a reporting export. The gap that matters is whether the system can be driven programmatically — whether an agent can change a routing rule, not just describe one.
- 50+
- skills and tools exposed through the MCP
- 5×
- faster territory planning
03Proof
What they say
We’ve automated our MQL distribution process with Gradient Works, which has been instrumental to scaling our sales organization for the next stage of our company’s growth.
For a two-person team like ours, we can now do things at a level and scale that would usually take a 5, 6, or 7-person RevOps organization.
04Questions
Questions we hear from GTM engineers
What can the agent actually change?
Routing flows, territory scenarios, overrides and deployments. Read-only access is the easy half; the useful part is an agent that can make the change and show you what it did.
Which agents are supported?
Anything that speaks MCP. Available today for Claude Desktop, Claude Code, ChatGPT, OpenAI Codex and Cursor.
Is routing logic inspectable?
Yes — everything lives in Salesforce Flow rather than an opaque external service, and Assignment Histories record every ownership change.
Do I need the managed package?
For routing and dynamic book management, yes. Scoring, enrichment and territory planning work against Salesforce, HubSpot or a CSV.
Put territory logic behind an API
Connect your agent to the MCP and start with the routing question that has cost you the most time.