GetProfile

GetProfile

User profiles and long-term memory for your AI agents

193 followers

Open-source drop-in LLM proxy that gives AI model persistent memory and structured user understanding.
This is the 2nd launch from GetProfile. View more

GetProfile

User profiles and long-term memory for your AI agents
Self-hosted and open-source user profiles and long-term memory for your AI agents. Unlike generic memory solutions that store blobs of text, GetProfile extracts natural language summary, typed traits with confidence scores and relevant memories with importance levels into a structured user profiles stored in your own PostgreSQL database.
GetProfile gallery image
GetProfile gallery image
GetProfile gallery image
GetProfile gallery image
GetProfile gallery image
Free
Launch Team
AppSignal
AppSignal
Get the APM insights you need without enterprise price tags.
Promoted

What do you think? …

Egor Miliukov
I'm excited to share our vision for the future of GetProfile AI as we embrace open-source. I believe that the best innovations happen when communities come together. That’s why I'm making the core components of GetProfile AI open-source. This move will allow developers worldwide to contribute, customize, and build upon it, fostering a vibrant ecosystem of innovation. What that actually means? Complete Transparency: You can now inspect the code, understand how it works, and ensure it meets your standards. Full Customization: Tailor the platform to fit your unique needs and use cases. Faster Innovation: With more minds working together, we can accelerate the pace of innovation and bring new features to life. Completely free: Self-host on your own infrastructure, no vendor lock-in.
Anton Loss

Congratulations on your launch! 🚀
On your home page, your example looks like something a agentic platform developer might use. Are they your prefect customers? Developers who build tools like Cursor and Claude Code? Or is this more suitable for individual developers? Or is it usable by Vibe Coders?

Egor Miliukov

@avloss Thank you!
Yes, developers who build agents and chatbots are the perfect customers.
Vibe Coders might find it hard to setup/deploy for now, probably, but I'm working on cloud version that will automate and simplify deployment, decreasing the barrier for adoption.

Jay Dev

Woah, GetProfile looks amazing! Persistent, structured user understanding is exactly what my AI projects need. Im curious, how well does the trait extraction handle nuanced or sarcastic language?

Egor Miliukov
@jaydev13 thanks! Glad you like the idea! Extraction quality very much depends on the used model and the prompt. Both are customizable, so it should be possible to configure the system to your specific requirements.
Germán Merlo
Sounds super interesting!
Rohith S

Congrats! May i know, what's the advantage of this over mem0?

Egor Miliukov

@rohith_s5 Thanks!

Mem0 and GetProfile solve the same core problem (LLMs don’t remember), but they “sit” in different places in your stack and push you toward different workflows.

Mem0 is a “memory layer” you typically integrate inside your app/agent code via their SDK: you add memories, search/retrieve relevant ones, and then build prompts using what was retrieved.

GetProfile is an OpenAI-compatible proxy (gateway): you route chat/completions through it (often by just changing baseURL + headers), and it automatically injects a profile summary/traits/relevant memories into prompts and updates that profile over time, no changes in your application required.

Alex Cloudstar

I’ve been duct-taping memory for a tiny support agent. Blobs get messy fast. Typed traits + confidence makes sense. Self-hosted + Postgres fits. Curious about PII handling/export. Saving this to swap into my weekend build.

BOS HUB

Congrats on the launch, can't wait to see how the project develops.

Right now, I am using IDEs and storing my project codebase docs in markdown files. Does this GetProfile help with that, our its more focused on another aspect?

Egor Miliukov
@bos_hub_us thanks! For your use case GetProfile won’t help. At least not yet 😉 I would recommend to look at Supermemory, it could be a better fit.