#LLMOps
5 related articles
Langfuse in Practice: Building an Observable Evaluation Platform for Harness Agents
Using a Harness multi-agent financial analysis project, this article explains the core concepts of the open-source LLMOps platform Langfuse: tracing, token cost and latency analysis, prompt version management, automated and human evaluation, and its capability boundaries—helping teams build enterprise-grade agent observability and evaluation systems.
The AI Coding Era: Replacing Manual Code Review with Evaluation Gates
When AI coding agents generate code faster than humans can review it, line-by-line review breaks down. Zoroval's evaluation gates use SHA-256 fingerprints, mandatory eval reports, and auto-expiring evidence to enforce quality gates for LLM behavior changes in CI.
Open-Source Alternatives to LangSmith: A Complete Roundup of 144 Self-Hostable LLMOps Tools
LangSmith's backend is closed source. This roundup covers 144 open-source, self-hostable LLMOps tools across 9 categories — observability, evaluation, prompt management, guardrails, and more.
LangSmith Practical Guide: Building Full LLMOps Workflows with Python
LangSmith is a LangChain LLMOps platform. This guide covers tracing, dataset evaluation, LLM-as-a-judge, Playground, Prompt management, annotation queues, monitoring, and a document extraction case study in Python.
Getting Started with Langfuse: A Deep Dive into the LLM Observability Platform
What is Langfuse? Explore this open-source LLMOps platform's core capabilities: tracing, evaluation, prompt management, and metrics — with support for LangGraph, LlamaIndex, and more.