Automated Multi-Modal Testing for Agents
AI agents often ship with undetected failures, inconsistencies, and poor response quality across real user journeys that are hard and time-consuming to test manually.
Ashr is an automated multimodal testing and evaluation platform that generates large volumes of realistic agent user journeys and custom metrics to detect errors and validate quality across tool calls, outputs, and interactions.
Teams building and deploying AI agents (developers and product/QA teams) who need automated end-to-end testing and evals.
built ai agents in finance and devops for the last two years. currently building ashr to work towards building better, more widely adopted agents.
Building AI powered workforce management software for small and medium sized businesses. Co-founder and CTO of Ask Geri (Acquired). Berkeley EECs.




