Back to browse
GitHub Repository

Pipevals is the visual pipeline builder for evaluation-driven AI development. Build evaluation graphs. Run them against datasets. Track quality over time.

14 starsTypeScript

Pipevals – a visual pipeline builder for evaluation-driven AI

by tilt·Mar 20, 2026·6 points·2 comments

AI Analysis

MidShip ItBold Bet

Early learning project in a crowded eval space dominated by LangSmith and Arize.

Strengths
  • Visual drag-and-drop canvas for composing evaluation graphs with typed step nodes
  • Human-in-the-loop review steps with configurable rubrics and multi-reviewer aggregation
  • Durable execution on Vercel Workflow with parallel branch support
Weaknesses
  • Author admits it's early and rough — zero stars, zero forks, learning project energy
  • LLM evaluation is saturated with LangSmith, Arize Phoenix, TruLens, and Braintrust
Category
Target Audience

ML engineers, AI product teams building evaluation workflows

Similar To

LangSmith · Arize Phoenix · Braintrust

Post Description

Hey HN! Pipevals is early and rough (this is a learning project), but usable.

It currently lets you: - build evaluation pipelines as graphs - run them against datasets - track how output quality changes over time

Similar Projects