Shared SaaS, Dedicated, or Your Own Data Center: How TotalGTM Deploys
TotalGTM runs three ways: managed multi-tenant SaaS, a single-tenant dedicated instance, or fully self-hosted in your own cloud or data center with local AI. Here's how to choose.
Most people meet TotalGTM as a SaaS product. You sign up, you log in, and you’re working. That’s the right starting point for most teams, and it’s how the majority of our customers run.
But “SaaS” is a delivery choice, not a fact of the software. The same TotalGTM that runs on our shared platform can run as a private instance we operate just for you, or entirely inside your own cloud account or data center. If you’ve ever had a security review stall a tool you wanted, this is the part worth reading.
Why deployment model matters
When your team researches accounts, builds ICPs, and drafts outreach, that work is sensitive. It maps out who you’re going after, what you know about them, and how you plan to win. For a lot of companies that data is fine in a managed cloud. For others, in regulated industries or with strict data-handling rules, it has to stay inside a boundary they control.
The usual answer is to pick between a great tool you can’t fully isolate, or an isolated setup with a worse tool. We didn’t want our customers stuck with that tradeoff, so the product was built to run in all three models without changing what it does.
The three models
Multi-tenant SaaS
This is the default, and for good reason. You sign up and you’re running the same day. We own the infrastructure, the updates, and the scaling. Every account is logically isolated, and you get new features the moment they ship.
It’s the right fit if you want to move fast and you don’t need a dedicated environment to satisfy a policy. Most teams start here and never need anything else.
Cloud-hosted dedicated
A step up in isolation. You get your own single-tenant instance with dedicated compute and dedicated data stores, on its own network boundary. We still run the operations, so it feels like SaaS to your team, but nothing is shared with another customer and you can pin your data to a specific region.
This is the common landing spot for companies that have data-residency requirements or an isolation mandate but don’t want to take on running the software themselves.
Your own cloud or data center
The full application deploys into your environment. Your data never crosses your boundary, and you hold the keys. This is the model for regulated industries, security-first teams, and anyone who needs to pass a hard review or run disconnected from the public internet.
Because the whole system is containerized and self-contained, this is genuinely the same product, not a stripped-down “on-prem edition.” You control access, encryption, logging, and backups. We work with your team to stand it up and support it.
The part that makes air-gapped real: local AI
Self-hosting your data only gets you so far if the product still has to phone an outside AI service to do its job. Every prompt would carry your data across the boundary you just worked to close.
TotalGTM’s AI layer is provider-agnostic. You can point it at the provider of your choice, or run a model locally so that inference happens entirely inside your network. In an air-gapped deployment, nothing leaves, not the data and not the prompts. There’s also a fallback path built in, so the work doesn’t stop if one provider has an outage. For a deeper look at why that fallback matters, the same thinking shows up in how we built the Claude MCP connector.
What about compliance?
We’re deliberate about this: a self-hosted deployment is built to fit your compliance program, not to hand you a checkbox. It runs inside your network boundary, writes to your logging, leaves your audit trail in your hands, and follows your data-retention rules. Rather than asking you to trust our environment, it puts the controls where your auditors already look.
If you have specific framework requirements, that’s a conversation we’d rather have directly than make claims about on a blog.
How to choose
A rough guide:
- If you want to start now and don’t have an isolation requirement, use the SaaS platform.
- If you need isolation or data residency but want it managed, go cloud-hosted dedicated.
- If your data and your AI both have to stay inside a boundary you own, go self-hosted.
You can also start on SaaS and move later. The product doesn’t change underneath you when you do.
If a dedicated or self-hosted deployment is on your radar, the security and deployment page lays out the models side by side, or you can talk to our team about your environment directly.