
2026-04-04 by Chris
Most chatbots work like forms in disguise. They ask a question, wait for your answer, then move on. It's a simple, linear, turn-based pattern. You'll find it everywhere. But it's also a fundamental limitation that rarely gets questioned.
Wanderer works differently. Wanderer chatbots are non-blocking, meaning the conversation keeps moving even while interactive elements are on screen. Buttons can appear, forms can be open, timers can be running, and the bot keeps talking, reacting, and adapting in parallel.
This isn't a workaround. It's a direct consequence of the underlying technology: Reactive Graph Sequencing (RGS).
In a traditional chatbot builder, every interactive element acts as a roadblock. When the bot shows a set of buttons, it stops. It waits. Nothing else happens until the user clicks. The entire conversation is paused by a single UI element.
For simple FAQ bots, that's fine. But as soon as you want something that responds to time, to external events, or to multiple things happening at once, the blocking model breaks down.
In a real conversation, the other person doesn't freeze mid-sentence just because they asked you a question. They might add context, react to your body language, or change the subject entirely. Traditional chatbots can't do any of this. Wanderer can.
The key is how Wanderer models conversations. Instead of a linear flow that moves from step A to step B to step C, Wanderer uses a reactive graph. Every element in the conversation (messages, buttons, input fields) is a node in that graph. These nodes don't wait for each other unless you explicitly tell them to.
In practice, that means:
None of these elements block the graph. They participate in it.
One of Wanderer's flow templates illustrates this well: a quiz game.
The bot asks a question and presents multiple-choice answers as buttons. A timer starts counting down. So far, nothing unusual. But here's where it gets interesting:
In a traditional chatbot builder, this would require custom code, complex workarounds, or simply wouldn't be possible. In Wanderer, it's just nodes in a graph.
The quiz is a useful demo, but the underlying principle has broader applications. A non-blocking chatbot can react to events that don't originate from the user.
A few examples:
None of this works in a blocking model. There, the bot is frozen, waiting for a click that might never come while conditions change around it.
So how does this work under the hood? The answer is Reactive Graph Sequencing (RGS), a paradigm that rethinks how chatbot logic is structured.
Traditional bots are event-driven: something happens, then the bot decides what to do next. RGS is state-driven: the graph continuously evaluates the current state of the entire conversation and determines what should be happening right now.
Concretely:
All of this happens visually, in the browser, without writing code. You connect nodes, define conditions using Must/May/Not edges, and the graph handles the rest.
Most chatbot builders treat non-blocking behavior as an edge case, something you patch in when the default flow isn't enough. In Wanderer, non-blocking is the default. It's not bolted on top; it's built into the architecture.
Every node is independent. Every node is reactive. Every node can respond to state changes from any source: user input, timers, external events, other nodes. The graph doesn't block because there's nothing to block. There's no single thread of execution waiting for a response. There's a graph that reacts to the world as it changes.
That's why Wanderer chatbots feel different. They don't wait, they don't freeze, and they don't leave stale buttons on the screen. They move with the conversation and with everything happening around it.
Wanderer is a visual chatbot builder powered by Reactive Graph Sequencing. Build non-blocking, reactive conversations without code. Learn more at wanderer-flow.de.
Title image: https://unsplash.com/de/fotos/ein-kleiner-wasserfall-mitten-im-wald-LTDMP9v21cc