Evidently Data Drift packaged by Software Sushi

This is a repackaged open source software product wherein additional charges apply for independent packaging, configuration and setup documentation. Software Sushi packages Apache-2.0 Evidently 0.7.23 with a pinned Python environment, automatic service startup, a local workspace and a controlled data-drift walkthrough. Software Sushi is independent of Evidently AI. Dependencies retain their own licenses. Evidently Cloud, upstream commercial support, managed alerting and hosted control-plane features are not included.

Compare reference and current Pandas datasets with Evidently Report and ValueDrift, retain reports in the native workspace, and export HTML and JSON. The supplied walkthrough creates two synthetic 200-row CSV datasets. It checks a deliberately shifted amount column and an unchanged stable column using two-sample Kolmogorov-Smirnov tests, independently rereads the CSVs, checks exact expected p-values and API report widgets, and saves a hash manifest. These controlled examples do not establish suitable thresholds or model quality for your data. No GPU, model download, LLM key or Evidently account is needed.

Launch t3.medium or larger x86_64 with at least 4 GiB RAM and 32 GiB root storage. Require IMDSv2 and encrypted storage. Connect with Systems Manager Session Manager or SSH as ubuntu, keep inbound application ports closed and forward local port 8000 through your authenticated tunnel. The native UI and API are loopback-only on port 8000. First boot creates a per-instance native write token, stored in a mode-0600 customer access file. The token protects writes; local users and people with tunnel access can read reports. This is a single-owner workspace without tenant isolation or per-user roles. Do not expose port 8000 publicly or share the tunnel with untrusted users.

Run sudo -iu ubuntu bash /opt/sushi/docs/demo.sh, then use your authenticated tunnel to inspect the created project, Reports tab and report. Read /opt/sushi/docs/README.md for Python examples, token handling, retention, backup and recovery. Projects and reports persist in /var/lib/evidently/workspace across service and instance restarts. There is no automatic expiry or external backup; deleting the root volume can delete reports. Monitor disk usage and encrypt backups because configuration backups contain the write token. EC2, EBS and networking charges are separate from the software price.

Software costs USD 0.03 per running instance-hour on each offered type. English getting-started guide: https://softwaresushi-marketplace-branding-public.s3.us-east-1.amazonaws.com/docs/evidently/63be81268680403f98c2/getting-started.html