Prompt management, LLM evaluations and observability
PromptLayer
PromptLayer builds an AI engineering platform combining a visual prompt CMS, evaluations and agent observability. Engineers and domain experts can edit, version, test and deploy prompts without redeploying application code. Teams can run dataset-based regression tests with human and automated graders, inspect agent tool calls, and connect production failures, costs and latency to specific prompt versions. Speak uses the platform for language-learning development, while Gorgias uses it to refine AI customer support.
Co-founders Jared Zoneraich and Jonathan Pedoeem met at high-school hackathons and initially built PromptLayer for their own AI projects. Zoneraich joined Cognition and announced that Pedoeem would take over as CEO, with Zoneraich staying on as an advisor. The platform’s registry maintains immutable versions, diffs and rollback, while SDK and REST API access let engineers integrate prompt retrieval and evaluation into software delivery workflows.
As of August 2026, the company reported 1,000+ companies using PromptLayer. By 2025, Gorgias had built a 60-person AI team around the platform, according to PromptLayer. PromptLayer raised a $4.8 million seed round in 2025, led by ScOp VC and Stellation Capital.
1 talk
Newest first
1 speaker at AIE
Affiliations reflect their AIE appearances, not necessarily current employment.
Messages from the stage
Testing coding-agent behavior
The workshop demonstrates historical backtesting, LLM assertions, and LLM-as-judge evaluation as techniques for assessing coding agents.
Affiliations reflect each recorded session, not necessarily current employment. The Claude Code workshop is an independent examination, explicitly not endorsed by Anthropic.