LLMs

Large Language Models (LLMs) are advanced AI systems trained to predict and generate human-like text based on vast amounts of data. These are constructed using transformer architectures and self-attention mechanisms to handle long-range context and complex language patterns effectively. LLMs are versitile, although the output is probabilistic rather than deterministic, responses should be evaluated to ensure accuracy and reliability.

LLMs are used in Planhats native AI features for summarizing, analyzing sentiment and classifying conversations, as well as in the Writing Assistant and AI Automation step.

Model alternatives: Managed LLMs vs Bring-your-Own

LLMs can be used in Planhat's Automations, to analyze and predict data. Planhat provides managed LLMs, which allows you to skip infrastructure setup and start generating outputs immediately with our out-of-the-box solution, we handle scalability and performance. You can choose between different models from providers like Google, OpenAI, and Anthropic, giving you the flexibility to select the most appropriate model for your specific task.

Alternatively, you can bring your own LLM from any provider, giving you full control over model selection, architecture, and fine-tuning. In this case, you're responsible for managing data residency, privacy, scaling, maintenance, and compute resources. Read about setting up your own external connection here.