fbpx
Logo
Logo
Introducing Controllable
Language Models™
Powered by neuro-symbolic reasoning, Controllable Language Models™ (CLM) outclass traditional LLMs in conversational AI, providing superior control, tool use and reasoning while evolving directly from human feedback. Explore Apollo, the platform to build CLM-powered conversational AI Agents that represent businesses.
About Controllable Language Models™
Developed over the past six years in collaboration with 60,000 human agents, Controllable Language Models™ (CLMs) are an evolution of traditional LLMs, unlocking a new phase in conversational AI. CLMs rely on neuro-symbolic reasoning rather than token prediction, resulting in superior reasoning capabilities to LLMs. They achieve a 100% success rate in tool use, never failing to execute API calls and always returning grounded answers. They learn directly from human feedback on each interaction, evolving from every conversation to align themselves better with their operators.

Instead of relying on a Large Language Model to orchestrate the entire conversation, CLMs classify each message to trigger the appropriate workflow, based on the current situation in the conversation and the customizable instructions of the agent. When the workflow is successfully completed (e.g. web search or API call), the CLM generates its response to the user. This method relies on a structured interaction state, achieved through sensory models that produce a symbolic, parameterized representation of each interaction.

About Apollo

Powered by our foundational CLM, Apollo-X, Apollo is a platform to build conversational AI Agents that can represent businesses. Overcoming the challenges that prevent LLMs from working on behalf of businesses, Apollo enables companies to responsibly communicate with their customers using conversational AI, and to sell their products and services without the need for human involvement.

To be able to represent companies responsibly and safely in front of their customers, Apollo offers fine-grained control by adhering to the deploying company’s policies, while providing a white-box view of its decision-making and reasoning. Apollo does not train on company data, instead, it can be pre-integrated with any number of internal systems, and call the relevant API endpoint in every message. Companies can improve their agents directly with human feedback right from the playground, resulting in ever-evolving agents.
Read more

Open Positions

Product Manager

New York
Learn More

Engineering Team Lead

Tel Aviv
Learn More

Product Manager

Tel Aviv
Learn More
See More Positions