Request Access

GA scheduled for December 2025

Technology

Product

Pricing

Company

Blog

Neuro-Symbolic AI

Terms of Use

Privacy Policy

Contact Us

Apollo-1 Playground

Apollo-1: The Foundation Model for Controllable Agents

The Control Problem

<hr>
<h3>01. The Control Problem</h3>
Two different things are emerging under the name &#8220;agent.&#8221;
Open-ended agents work for users. Coding assistants, computer-use agents, personal AI. You&#8217;re the principal—the one whose goals matter. If the agent interprets your intent slightly differently each time, that&#8217;s fine. You&#8217;re in the loop. You&#8217;ll correct it. Flexibility is the point.
Task-oriented agents work on behalf of entities. An airline&#8217;s booking agent. A bank&#8217;s support agent. An insurer&#8217;s claims agent. These agents serve users, but they represent the entity. The entity is the principal—the one whose policies must be enforced. The agent has to follow the entity&#8217;s rules while conversing naturally with customers.
The requirements are fundamentally different.
Open-ended agents need maximum flexibility. They should handle novel situations, figure things out, do whatever the user wants. LLMs are well-suited here—probabilistic, creative, adaptable.
Task-oriented agents need control. Specific behaviors must execute in specific scenarios. &#8220;If the refund exceeds $200, always require ID.&#8221; &#8220;Always offer insurance before payment.&#8221; These aren&#8217;t suggestions. They&#8217;re requirements that determine whether AI can be trusted with customer interactions involving real money, real appointments, and real business logic.
LLMs can&#8217;t guarantee these behaviors. Not because they&#8217;re not smart enough, because they&#8217;re architecturally incapable of it. They predict tokens. They approximate intent. They do what you want most of the time.
Most of the time isn&#8217;t good enough when the stakes are real.
Two kinds of agents. Two different foundations required.

Why LLMs Agents Can't Solve It

<hr />
<h3>02. Why LLMs Agents Can&#8217;t Solve the Control Problem</h3>
The opportunity is enormous. Every conversation that results in real-world action—booking flights, processing payments, managing claims, executing trades—could be automated. These interactions run the economy. The market for task-oriented agents dwarfs what open-ended assistants will ever capture.
But without control, task-oriented agents can&#8217;t be deployed reliably at scale. Enterprises won&#8217;t trust AI with customer interactions when &#8220;usually works&#8221; is the best guarantee available.
The industry has spent three years and billions trying to make LLMs agents work. The approaches vary. The tradeoff doesn&#8217;t.
Prompting: Tell the model to &#8220;always do X&#8221; and hope it complies. You get fluency. You don&#8217;t get control.
Fine-tuning: Train the model on task-oriented behavior. It learns patterns. Patterns improve probability. They don&#8217;t create guarantees.
Orchestration: Wrap the model in workflow frameworks. Build state machines around it. Route between prompts. You get control over the scaffolding, but the LLM inside the wrapper is still probabilistic. The moment a user goes off-script, the system either breaks or falls back to uncontrolled generation.

What Controllable Agents Need

<hr />
<h3>03. Controllable Agents</h3>
Task-oriented agents need control. But &#8220;task-oriented&#8221; describes the use case, not the capability. Most task-oriented agents today are just LLMs pointed at tasks. They inherit all the limitations we just described.
We need a term for agents that actually deliver what task-oriented use cases require: Controllable Agents: agents whose behavior can be programmed in critical scenarios while preserving natural conversation everywhere else.
Controllable Agents are a new class. Not a product category or a marketing term but rather a computational category, defined by capabilities that LLMs structurally cannot provide.
What would a foundation model for Controllable Agents require?
Explicit state. Multi-turn interactions require tracking where you are in a process, what you know, what&#8217;s happened, what constraints apply. Language models have no native state, they rely on context windows and external memory.
Programmable behavior. Entities need to define how their agent behaves in high-stakes scenarios and know those definitions will hold. &#8220;If X, always do Y.&#8221; Not probably. Always. Language models are probabilistic by design.
Native tool use. Controllable agents invoke external systems: booking engines, payment processors, CRMs. These invocations need reliability, proper parameters, error handling, execution guarantees. Language models sample tool calls probabilistically.
White-box reasoning. Every decision must be traceable: which rules fired, how state evolved, why the agent acted. Language models are black boxes.
Fluent interaction. Despite all of the above, the agent still needs to converse naturally, handle unexpected inputs, respond like humans expect. Rigidity kills user experience.
This is the core tension: you need the fluency of neural language generation and the reliability of symbolic control. You need both, unified—not bolted together.
Controllable Agents need their own foundation.

Apollo-1

<hr />
<h3>04. Apollo-1</h3>
Apollo-1 is the first foundation model built for controllable agents.
Not a language model adapted for control. Not an orchestration layer around existing models. A new foundation, built from the ground up on neuro-symbolic architecture that unifies generation and control.
Apollo-1 combines neural modules that understand and generate natural language with symbolic modules that maintain state, enforce rules, and guarantee execution. The neural components interpret meaning, handle ambiguity, produce fluent responses. The symbolic components track state, apply logic, ensure defined behaviors execute exactly as specified.
The agent understands language like an LLM. It enforces behavior like a state machine. One model. Both capabilities. Native to the architecture.
You don&#8217;t program every interaction, only the ones that matter. For everything else, the agent defaults to intelligent, fluent, common-sense conversation. When you&#8217;ve defined specific behaviors, they execute with certainty. When you haven&#8217;t, the agent thinks for itself.

<hr />
<h3>01. The Control Problem</h3>
Two different things are emerging under the name &#8220;agent.&#8221;
&nbsp;
Task-oriented agents work on behalf of entities. An airline&#8217;s booking agent. A bank&#8217;s support agent. An insurer&#8217;s claims agent. These agents serve users, but they represent the entity. The entity is the principal—the one whose policies must be enforced. The agent has to follow the entity&#8217;s rules while conversing naturally with customers.
The requirements are fundamentally different.
Open-ended agents need maximum flexibility. They should handle novel situations, figure things out, do whatever the user wants. LLMs are well-suited here—probabilistic, creative, adaptable.
Task-oriented agents need control. Specific behaviors must execute in specific scenarios. &#8220;If the refund exceeds $200, always require ID.&#8221; &#8220;Always offer insurance before payment.&#8221; These aren&#8217;t suggestions. They&#8217;re requirements that determine whether AI can be trusted with customer interactions involving real money, real appointments, and real business logic.
LLMs can&#8217;t guarantee these behaviors. Not because they&#8217;re not smart enough, because they&#8217;re architecturally incapable of it. They predict tokens. They approximate intent. They do what you want most of the time.
Most of the time isn&#8217;t good enough when the stakes are real.
Two kinds of agents. Two different foundations required.

<hr />
The opportunity is enormous. Every conversation that results in real-world action—booking flights, processing payments, managing claims, executing trades—could be automated. These interactions run the economy. The market for task-oriented agents dwarfs what open-ended assistants will ever capture.
But without control, task-oriented agents can&#8217;t be deployed reliably at scale. Enterprises won&#8217;t trust AI with customer interactions when &#8220;usually works&#8221; is the best guarantee available.
The industry has spent three years and billions trying to make LLMs agents work. The approaches vary. The tradeoff doesn&#8217;t.
Prompting: Tell the model to &#8220;always do X&#8221; and hope it complies. You get fluency. You don&#8217;t get control.
Fine-tuning: Train the model on task-oriented behavior. It learns patterns. Patterns improve probability. They don&#8217;t create guarantees.
Orchestration: Wrap the model in workflow frameworks. Build state machines around it. Route between prompts. You get control over the scaffolding, but the LLM inside the wrapper is still probabilistic. The moment a user goes off-script, the system either breaks or falls back to uncontrolled generation.

Accordion title

Gemini

Steps

<hr />
<h3>04. Apollo-1 TES TES TE</h3>
Apollo-1 is the first foundation model built for controllable agents.
Not a language model adapted for control. Not an orchestration layer around existing models. A new foundation, built from the ground up on neuro-symbolic architecture that unifies generation and control.
Apollo-1 combines neural modules that understand and generate natural language with symbolic modules that maintain state, enforce rules, and guarantee execution. The neural components interpret meaning, handle ambiguity, produce fluent responses. The symbolic components track state, apply logic, ensure defined behaviors execute exactly as specified.
The agent understands language like an LLM. It enforces behavior like a state machine. One model. Both capabilities. Native to the architecture.
You don&#8217;t program every interaction, only the ones that matter. For everything else, the agent defaults to intelligent, fluent, common-sense conversation. When you&#8217;ve defined specific behaviors, they execute with certainty. When you haven&#8217;t, the agent thinks for itself.