At Quess Corp (Ethara.ai), I build AI evaluation harnesses and multi-turn response trajectories to guide model outputs toward precise, consistent goals. I also develop rubric-based grading frameworks and feedback loops to evaluate model performance and support alignment.
I benchmark model outputs across structured datasets to identify failures and optimize prompts. I deploy workloads using AWS ECR and EC2, and use Git in collaborative release workflows.
Previously, at Homenetics Technology Pvt Ltd., I developed full-stack applications and Eagle Eye, a device monitoring tool with status tracking, restart functionality, and error reporting. I also configured WireGuard VPN across production servers and integrated Google APIs into the Food Explorer platform.
