At Turing, I fine-tune Claude agents against OSWorld benchmarks and build Python evaluation pipelines to measure task correctness, safety, and instruction-following. I also prepare SFT datasets and benchmark tasks with validators and structured logs.
As a Freelance Engineer, I created Python data-processing utilities and a Bash monitoring tool that sends email alerts for website performance. I also developed Django dashboards and automated network configuration backups.
At Rakuten Mobile and Tech Mahindra, I worked on Python automation, network configuration management, and deployment workflows. My work included Ansible playbooks, CI/CD pipelines for Python services on AWS EC2, and automated pre- and post-change network validation.
