Log inSign up
Log inSign up
GlassFlow
127 posts
GlassFlow profile banner
@glassflowdev

GlassFlow

@glassflowdev
GlassFlow builds the data infra for AI agents in production. Tares feeds agents correlated data from every system & Rius traces and debugs what prod agents do.
Berlin (Germany)
glassflow.ai
Joined October 2023
11 Following
60 Followers
RepliesRepliesRepostsRepostsMediaMedia

Log in or sign up for X

See what’s happening and join the conversation

Continue with phone
or
Log in with username or email
Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
  • @glassflowdev
    GlassFlow
    @glassflowdev
    15h
    Your agent finished the run. No error or alert. Well... it was wrong Here’s how to catch silent failures mid-run: ➡️ a 5-stage framework ➡️ the signals to watch ➡️ a setup you can ship in a week See more: glassflow.ai/blog/how-to-de… #AgenticAI #AIAgents
    How to detect AI Agent failures in production
  • @glassflowdev
    GlassFlow
    @glassflowdev
    Oct 5
    10 days to go ⏳ We're in London on Oct 15 for the next #AI Signals meetup, hosted by @Zoopla. We will show how to catch silent agent failures mid-run with a live look at #Rius. Spots are limited, so reserve yours today 👉 luma.com/cpyel4un #AIAgents #Observability #Agentic
    AI Signals #35: Evals, Guardrails and Human Behaviour · Luma
    From luma.com
  • @glassflowdev
    GlassFlow
    @glassflowdev
    Oct 2
    Your AI agent looks healthy. But what about your customers? New in Rius: the Users view. Per-user traces, failures, error rate, cost and TTFT p50. Customer says "it didn't work"? Find the user. Open their run. Follow the failure. docs.glassflow.ai/rius/platform/…
    1
  • @glassflowdev
    GlassFlow
    @glassflowdev
    Sep 29
    Rius traces your AI Agents live. Not after they crash. Observability for agents in production, from $19/month. Try today for free: glassflow.ai/rius?utm_sourc… #AgentObservability #LLMtracing #AIEngineering #AIAgents
    00:00
  • @glassflowdev
    GlassFlow
    @glassflowdev
    Sep 21
    An agent that crashes is the easy case: you get a stack trace. The hard case? It finishes every step, returns a confident answer, but it's wrong. 6/ 7 ways AI agents fail in production raise no error. glassflow.ai/blog/ai-agent-… #AIAgents #LLMOps #AgentObservability
    The 7 ways AI Agents fail in production
    2