How Snorkel AI uses Portkey
2x
“2x faster issue detection in development”
As published on portkey.ai. Captured by usedby on Oct 5, 2026.
What happened
Snorkel AI uses Portkey to trace and debug its multi-agent evaluation systems that validate expert-generated labels. Portkey shows every agent, sub-agent, and tool call as a nested tree, and lets the team compare good and bad traces side by side.
Summary written by usedby from the source page, in English. The figures are those of Portkey and Snorkel AI, not ours.
- 20%“20% accuracy boost in agent evaluations by catching and fixing edge cases that were previously invisible in aggregate metrics.”
With Portkey, the team could finally see what was happening inside their agents. Every call, every sub-agent, every tool invocation appeared as a nested tree, no more reconstructing execution paths from scattered logs.
Prior to Portkey, Snorkel debugged LLM applications with a fragmented patchwork of logs in s3, DB queries, and manual visualization. The clean Portkey UI allows us to identify, debug, and rebuild agents with confidence.
With Portkey, every agent, sub-agent, and tool call is visible in one hierarchy — you can see how decisions branch and converge. Seeing the full execution tree in one view completely changed how we debug
What the story claims, and what we checked
We compared the story with its live page on Oct 5, 2026.
- The figure: 2xCheckedPrinted word for word on the page, near the name of Snorkel AI.
- The passage quoted aboveCheckedCopied word for word from the page, near the name of Snorkel AI.
- Snorkel AI uses PortkeyCheckedConfirmed line. Latest check across sources: Oct 5, 2026.
- The result itselfNot checkedWe quote it; we did not measure it.
More stories
Same company, same tool or same industry.
