Langfuse

Langfuse is now much more accessible for Agents: try the new CLI, skills, MCP server, and RAG api. Start with the skill as it helps agents onboard onto all of these features.

We test some of Langfuse v4 (GA soon) on Langfuse Cloud and the feedback is crazy Thankful to work with a product engineering team (Steffen, Valeriy, Nimar, Marlies, Hassieb, Max) that cares so deeply about pushing these metrics and for having joined the best OLAP db team at ClickHouse At this scale, we are effectively a systems/db company

Now live: Langfuse x ClaudeCode integration. Use it to capture full conversation interactions. See how your team is using ClaudeCode. Set it up by telling Claude to look at https://lnkd.in/dr23VBhz

Langfuse x LibreChat integration is now live. Shoutout to Danny Avila for the ship!

We're hiring our first product designer 🎨 The challenge: LLM observability tools are inherently data-dense. Traces, evals, prompts, metrics—lots of information engineers need to debug and improve their AI apps. Your job is to make this complexity feel simple. We're looking for someone with great taste who is technical and obsesses over developer UX. You'll own the interface end-to-end: simplifying complex workflows, designing filters that actually work, and making Langfuse the best devtool UI in the space. You'll talk to users a lot, and shape how thousands of teams interact with their LLM applications daily. Check out the job description and reach out if this sounds like your calling -> ordnl.link/fH2xFyJ

Langfuse joins ClickHouse! We started Langfuse in 2023 to solve a problem we kept hitting ourselves: building LLM apps is easy to demo, hard to run in production. Thousands of teams now use Langfuse to trace, evaluate, and iterate on their AI applications. As we scaled, we moved our data layer to ClickHouse, and that technical decision turned into a real partnership. We collaborated on features, shared customers, ran community meetups together. ClickHouse uses Langfuse internally; we're a large ClickHouse Cloud customer. Building Langfuse with the team is so much fun. If it was not for long conversations with Aaron, Yury, Alexey, Tanya, Ryadh, and Pete, realizing how much we can accelerate together, and the OSS and engineering focused culture that is super aligned, I would have never considered this path. Extremely excited to build together! Why we decided to join - Langfuse is a data-intensive product. Working closely with ClickHouse engineering helps us push performance and reliability faster than we could alone. - ClickHouse has every incentive to keep Langfuse fast and dependable at scale; we share customers, OSS deployments, and now a roadmap. - The culture is a match: open-source identity, developer-first instincts, performance as a product feature. What stays the same: - Ambition to build the best product possible to accelerate AI Agent deployment - Open source and self-hostable (no licensing changes planned) - Langfuse Cloud keeps running as-is - Team joins to keep building Langfuse and we will expand the team a lot Grateful for the community and team that got us here. This lets us put more resources behind building something we all care deeply about! Congrats all, Marlies, Hassieb, Jannik, Steffen, Felix, Nimar, Akio, Valeriy, Lotte, Leonard, Max and Clemens, the job is not finished, but today is a great day to celebrate! -> Blog post with full story in comments

So excited to finally drop this. Langfuse joins ClickHouse to push Agentic AI development forward. Open Source. Faster than ever. Full story in the comments.

Curious how the best high-velocity teams run Product Engineering? We’re hosting a meetup with our friends at Tanso and Center for Digital Technology and Management (CDTM) to unpack the mechanics of the modern tech stack: merging Eng, PM, and Design to ship faster than ever. Join us for an evening of showcases and open discussion. Whether you are a (Product) Engineer, a Founder, or just curious about the shift—block your calendar and sign up here: https://luma.com/k2vmo9gh. Capacity is 80% full.

⭐ 20,000 stars on GitHub! ⭐ It took us 2 years to reach 10,000. 8 months later, we doubled that! We couldn't build Langfuse without the most amazing community! Thank you!

Langfuse Context: How Superhuman built Agents that are used by millions (Jan 15) Together with our friends at Superhuman, we're hosting an exciting event next week in Berlin. This time our CEO Marc Klingen is joined by Ailian Gan (Director of AI Product at Superhuman). Sign-up link below, seats limited. Ailian and Marc will chat about how Superhuman builds and scales AI agents used by millions daily, covering their product team structure, how they identify LLM use cases that users actually need, and designing interfaces for mass-market adoption. The chat will be followed by drinks and food with lots of space for networking and discussions. Sign up here -> https://luma.com/lahwzvtt