Worked on enhancing the stability of single-node deployments in the apache/texera repository by updating the Docker Compose restart policy. The approach involved changing the policy from 'always' to 'unless-stopped' using YAML configuration, ensuring that services automatically recover from failures while still respecting user-initiated stops. This adjustment improved uptime and gave users greater operational control, with one-shot jobs remaining unaffected by the change. Demonstrated DevOps and containerization expertise through careful policy selection and precise code commits. The work focused on practical reliability improvements, leveraging Docker and YAML to address deployment resilience without introducing unnecessary complexity or affecting existing workflows.