3.3.4 LLM Integration

Orbit Insight leverages advanced Large Language Models (LLMs) to power its data analysis, search, and chat capabilities. Integrating with these models enables the platform to deliver high-quality insights, accurate information retrieval, and sophisticated natural language processing.

1. Production Integration with OpenAI’s Latest Model

In its production environment, Orbit Insight is exclusively integrated with OpenAI's latest model, currently ChatGPT-4o. This model is chosen for several key reasons:

• Advanced Capabilities: ChatGPT-4o remains one of the most advanced LLMs available today, offering unparalleled performance in natural language understanding and generation. This ensures that Orbit Insight can deliver the most accurate and contextually relevant insights to its users.

• Scalability and Support: OpenAI provides robust production-level support, particularly in terms of scalability. As Orbit Insight processes large volumes of data and supports numerous users, it’s critical to rely on a model that can scale effectively without compromising performance. OpenAI’s infrastructure is designed to handle such demands, ensuring consistent and reliable service.

• Continuous Updates: By integrating with OpenAI’s latest model, Orbit Insight benefits from continuous updates and improvements. As OpenAI releases new features and enhancements, they are automatically incorporated into the platform, keeping Orbit Insight at the forefront of LLM technology.

2. Addressing Data Security and Compliance Concerns

While the integration with OpenAI’s model provides significant advantages, Orbit Insight recognizes that some clients may have specific data security and compliance concerns. To address these concerns, Orbit offers the flexibility to deploy alternative, open-source LLMs as requested by clients.

For clients who require additional control over their data and infrastructure, Orbit Insight can deploy open-source LLMs within their private environments. This option provides greater oversight and compliance with internal data security policies, ensuring that sensitive information is handled according to the client’s requirements.

Detailed information about these deployment options, including how open-source models can be integrated and managed within a client’s infrastructure, can be found in the Deployment Options chapter. This section covers the various configurations and support provided to ensure a seamless and secure integration of LLMs, tailored to the client’s specific needs.