Understanding BotSpot

Hi — BotSpot is basically a platform built on a few principles.

The first one is that we want to build it as a governance-first platform. Governance isn't a feature we bolted on, it's a base pillar. The problem it solves: enterprise customers, and normal SMB customers too, already forget that they need privacy. It's their right, and it's a responsibility — and they're not educated about it, so they forget. That's our first principle.

The second principle is that we have a three-surface system. A developer surface called Workbench, a distribution surface called Workforce, and a consumption surface called MyCrew. On Workbench, people build AIs — applications, agents, anything an AI could do. That's the build surface. If you're a marketer or a publisher with a really good AI agent you'd like to bundle and sell in a customizable way — think of it like an e-commerce site for agents — that's what Workforce is for. And the third layer is MyCrew, which is like WhatsApp and the other tools you already use: you just chat with agents, or get work done by one. This is where human–AI interaction is going to go deep. We split it into these three surfaces on purpose. That's our second principle.

These two are the primary principles.

Our third most important principle is that we always want to be pay-per-use. The token economy runs as a pay-per-use system. Some people want to build and host their own big LLMs, sure — but if you can consume an LLM at very low cost, most businesses would rather do that. If a centralized hosting system can guarantee their data privacy, people would love to use it and just get the job done. It's really only big enterprises that think about owning an LLM. Everyone else just wants to use one. For those people, we always want to give a pay-per-use surface. BotSpot Cloud — the .app — is always going to be a pay-per-use system. We're building every way and every means to calculate the use and pass it to the customer, so every turn, every chat, is like a debitable wallet. We don't want to give a monthly subscription model, because that's like a gym membership — you pay, and you don't go. That brings a problem, and that's exactly what a subscription is.

We've built BotSpot as a system where everything revolves around AI as the core, central piece. That's the design thinking behind it. Everything we do differentiates this product from the rest — even workflow orchestration is centered on the needs of a futuristic system that's completely dependent on AI-based work. For every AI-based piece of work, instruction is primary. Keeping that as the starting fact, we start with Prompt Center, where prompts are saved, versioned, governed.

Now that we've looked at the prompt as one tenant, the next thing to look at is the brain of the whole system — the models. Today's generative models handle multiple data types — text, images, video — and probably tomorrow audio too, if it isn't already. We're also thinking about telephony, which is a layer of its own, because that's how customers want to see it. We're bringing all of this inside BotSpot. So models are our second tenant. Prompt as a tenant is one decision; models as a second tenant is another.

The next is a factor everybody loves — the no-code factor. No-code means you drag and drop your requirements as basic primitives and derivatives, and build everything you need as a workflow, instead of typing or coding it out. That's our third tenant: workflows. We didn't pick this because it was the founder's personal choice — it's a populist, popular choice.

The other layer we're currently building, one that keeps evolving, is the data tenant. Data is a place where people want to see, interact with, and play around with their own data — not just an import-and-run-away model. It's interact, play, and model with data as well. That's the fourth tenant.

The last tenant is tools. Why tools? Because it's the most important way AIs are evolving — they're not just classifiers anymore. They're used to interact with other applications, private deployments, and internal data exposed as tools. And MCPs, which are a way to access other applications. Tools is the playground where people bring MCPs they already know, APIs they already know, or build their own Python tools or REST connectors. All of that is handled here — the tools tenant.

Finally — if I have to give you an idea of how to work with BotSpot, it's the interaction of all five of these. The best example is the Apps zone we've built, because people want to build their own software — these five are your core zones. You'll need a backend, a frontend, and your data. And the way today's systems are evolving is simple: your backend is your workflow, your frontend is built by AI agents, and your data is your database handling your data. That's a testament to how BotSpot is going to keep evolving. Apps isn't a tenant by itself — it's an amalgamation zone of all the basic tenants, and how they interact and play together. BotSpot gives you the full freedom, as a developer, to build anything. And everything.