Nebuly provides SDKs that make integration with your LLMs straightforward. Alternatively, you can use our API to send interactions. Nebuly extracts conversation logs and associated metadata asynchronously, i.e. with no impact to latency.Nebuly supports both open-source LLMs and API models, including:

• OpenAI
• Azure
• Anthropic
• Hugging Face
• AWS Bedrock
• Google


Please, visit Nebuly Docs to read more.

Nebuly offers two unique benefits for companies developing LLM-powered products.

1. Enhance your AI product experience.
Nebuly automatically and accurately captures implicit feedback  from AI-user interactions. Our customers have reported a 100x increase in actionable feedback from these interactions. Additionally, Nebuly's tools enable seamless integration of this feedback back into your product, driving continuous improvement.
‍
2. Unlock valuable business insights.
LLM interactions are a goldmine of insights. Nebuly identifies and highlights the most important topics and business insights directly from user interactions. For example, our customers have been able to pinpoint gaps in product offerings and uncover emotional associations with certain services, leading to targeted adjustments that improve conversions and drive revenue growth.

Nebuly is often considered part of the broader LLM monitoring stack because of its ability to monitor user interactions with LLMs. However, unlike traditional observability solutions Nebuly does not focus on model-level metrics (accuracy, recall, traces, etc) but on LLM users behavior. In fact, Nebuly tracks user satisfaction, sentiment, and conversation topics, highlighting issues such as unhelpful responses or interactions with malicious intent.

Nebuly is a user analytics solution built for the AI era. Unlike traditional product analytics tools that focus on point-and-click interfaces, Nebuly analyzes conversational AI interactions. It provides product managers with deep insights from conversations, such as poor responses from the LLM and emotional associations with specific conversation topics.