GetProfile is a self-hosted and open-source solution for managing user profiles and long-term memory for AI agents. Designed to enhance the interaction between AI models and users, it provides structured, context-aware data that can be injected into prompts without overwhelming the model with unstructured text. Unlike generic memory solutions, GetProfile extracts natural language summaries, typed traits with confidence scores, and relevant memories with importance levels, storing them in a PostgreSQL database for persistent and scalable use.
The product offers an OpenAI-compatible proxy, allowing seamless integration with existing AI workflows. This means developers can enhance their applications with personalized and context-aware interactions without modifying their existing codebase. GetProfile also enables full customization of user traits, giving developers control over what data is collected, extracted, and injected into prompts. It is designed for ease of deployment using Docker, making it simple to integrate into any environment.
GetProfile operates as a proxy between AI agents and LLM providers. Here’s how it works:
This process ensures that each interaction benefits from the user's history and traits without increasing the complexity of the prompt or affecting the model's performance.
| Benefit | Description |
|---|---|
| Personalized Interactions | Injects user-specific information into prompts for more tailored responses |
| Improved Context Awareness | Maintains relevant memory and traits across interactions |
| Enhanced Privacy | Data stays with the user, hosted in their own PostgreSQL database |
| Seamless Integration | Works with existing OpenAI-based systems without code changes |
| Scalable Architecture | Deployed via Docker, easily scaled with Docker Compose |
| Transparent and Auditable | Open-source code allows for review and customization |
GetProfile is ideal for developers building AI-powered applications that require persistent, structured user understanding—such as chatbots, virtual assistants, and personalized recommendation systems.
