
Open source AI engineering platform. Now part of @clickhousedb. We're hiring: https://teapawt.net/t.co/k6dgv4dws2
github.com/langfuse/langfuse
Joined July 2023
- Tweets1.1K
- Following682
- Followers6.8K
- Likes4K
Pinned Tweet
Replying to @marcklingen
@marcklingen sat down to demo all langfuse features in a single video.
Agent observability, monitoring, evals, prompt management, and what is special about Langfuse.
Great place to start if you are new, or catch up on latest changes.
langfuse.com/watch-demo
chat agents go beyond single turns with user questions and agent reply
understanding how the agent is used and how it fails requires following the overall flow of the interaction
follow interactions like users experienced them with the sessions view
dive deep into the sections where things went wrong
langfuse.com retweeted
you know the team is strong when the releases look like this multiple times A DAY
langfuse.com retweeted
Join me for a live lecture on this tomorrow 10am PT! thegenacademy.com/lesson/opt…
Editing the last slides, really enjoying how simple and clear some of them are
teams in regulated industries can’t ship an agent to production without evals
@annabellschfr on what's to learn from how they do it
66% of customer service orgs already use AI agents, according to Salesforce’s State of Service.
the key advantages: instant answers around the clock, and a big share of the workload off support teams
working with Langfuse customers running AI-powered support in production, here’s what we’ve seen work
a selection of customers running support chat agents in production
langfuse.com/users/canva
langfuse.com/users/sumup
langfuse.com/users/dkb
langfuse.com/users/cresta
we are a launch partner in the @FireworksAI_HQ Frontier Ecosystem!
running open models on Fireworks? Langfuse shows how they perform in prod, so you can improve your apps and turn real traffic into better, cheaper fine-tuned models.
teapawt.net/FireworksAI_HQ/status/…
The future of AI is specialized intelligence with models tuned to your data, your domain and your customers.
Today, we're proud to launch live from Forge the Fireworks Frontier Ecosystem with 60+ partners across the AI stack, helping you build and ship faster.
Check it out: fireworks.ai/partners
Skill management is now available in Langfuse, stay tuned for more 👀
You can now store Skills in Langfuse!
This means also detailed skill analytics on your traces.
Which skill was used how much?
Which files in the skill was actually read? Why?
and a lot more! Will share more over the coming weeks.
see: langfuse.com/changelog/2026-…
big ship by Tobi!
unhappy, frustrated and disagreeing users are the best signal something is going wrong in your chat agents
learn how to catch those signals on your traces in langfuse
your users usually show you where things are going wrong in your chat agent
by being frustrated, by asking follow up questions, by disagreeing with your agent response
check out how you can catch conversation signals in your chat agent traces in @langfuse
we templated how we observe and improve our own docs chat agent:
- trace your agent across turns
- find the interesting traces in production by catching conversation signals
- turn them into repeatable cases
- evaluate them to make sure you're improving
langfuse.com/blog/2026-07-16…
check out how we trace our docs chat agent into langfuse
every step on record, full view on cost and latency, turns grouped into sessions to follow the flow like our users did
build chat agents you can trust in production
Langfuse helps you trace and evaluate every session end to end:
- see what each turn cost
- find where conversations go sideways
- measure and improve quality with evals on real cases
Works with any model and any framework.
langfuse.com/chat-agents
RT @marliessophie: We 100x our team in 2 years.
> Then: @langfuse offsite 2024 with 7 people
> Now: @ClickHouseDB offsite 2026 with 800 p…
Two Langfuse plugins are now available on the @SpaceXSI Marketplace:
→ 𝘓𝘢𝘯𝘨𝘧𝘶𝘴𝘦 𝘖𝘣𝘴𝘦𝘳𝘷𝘢𝘣𝘪𝘭𝘪𝘵𝘺 captures your agents’ model calls, tool executions, and token usage.
→ 𝘓𝘢𝘯𝘨𝘧𝘶𝘴𝘦 gives your agents skills to investigate traces, manage prompts, and run evals.
Analytics powered by @ClickHouseDB
“List this week’s most expensive Cursor turns and what they have in common.”
Ask Cursor, update your instructions, then compare cost, latency, and task completion.
Read the guide: clickhouse.com/blog/cursor-a…
langfuse.com retweeted
Ask Halo challenge
The brief: find out why eligible customer applications were getting rejected.
In seven minutes, @langfuse's Halo:
- found an ignored €6,000 subsidy in traces
- tested a revised prompt and
- made it live, fixing the application workflow
Here's the demo.
New article on how we built Halo 😇
A couple of decisions:
- Halo uses the same MCP server, skill and docs as external agents accessing Langfuse, so every new Halo functionality is immediately available to everyone.
- The agent loop runs separate from its sandbox.
- We use Langfuse to observe and improve Halo.
Read the full article here: langfuse.com/blog/2026-10-21…















