LiveLangfuse · AI observability

Langfuse: trace detective

A user complains. You get the trace. Tap the step that caused it — six cases, three lives.

Case 1 / 6♥♥♥0 pts

“I waited 9 seconds for a simple answer.”

Tap a step to look inside it. Then blame the one that caused this.

Trace · support-chat9.2 s

Made-up traces, drawn the way Langfuse shows them: nested steps on a timeline, with model calls, tools, tokens, cost, prompt versions and scores.

How it works

1Langfuse records every run of an AI app as a trace: each step, how long it took, the tokens and cost of each model call, the prompt version and any scores.
2Each case gives you a complaint (slow, wrong, expensive, wrong language, stuck, 👎) and one trace. Find the step that caused it.
3Every case ends with the one habit it teaches — like “slow? read the timeline first”.
4The traces are made up, and drawn the way Langfuse shows them.

Read the post: LangGraph and Langfuse, in plain words

More demos

All demos

LiveDatabases · Built from scratch

QueryLite: a SQL database

A SQL database written from scratch — parser, B+ tree indexes, a query planner, joins and transactions — running live on 22,000 rows in your browser.

Try it

LiveDistributed systems · Built from scratch

Raft consensus, live

The algorithm that keeps etcd and Kubernetes consistent. Five servers elect a leader and replicate a log — crash them and split the network while it runs.

Try it

LiveReal-time · Built from scratch

Live collaborative editor

Real-time editing with no server. Three devices share a note — take one offline, edit everywhere, reconnect, and they merge. Open a second tab and it syncs live.

Try it