Upstream

Upstream includes five core capabilities:

• Dynamic Config API: A REST API for flags with multi-environment support, audience targeting, and percentage rollouts, with edge latency under 10ms.
• AI Rollout Pilot: Define SLOs and Upstream automatically ramps up the rollout percentage when metrics are healthy, or pauses if they degrade.
• Anomaly Guard: Real-time anomaly detection on business and technical signals, with auto-rollback or human-review pause based on severity.
• Codebase Intelligence: SDK introspects your codebase, maps flag usage, detects dead flags (unused for over 30 days), and auto-generates docs.
• Config Copilot: Ask questions in plain English like "which flags are active for Enterprise customers in EU?" and get answers with 1-click actions.

Upstream is used by engineering teams to safely roll out features. For example, a staff SRE at a fintech company had a flag rollout spike error rates at 2:47am; Upstream caught it and rolled back within 12 seconds, and the SRE found out the next morning via Slack. Another team migrated 80 flags from LaunchDarkly in one afternoon and used AI Rollout Pilot to replace 90% of manual ramp-up work. The Codebase Intelligence feature helped a VP Engineering find 14 dead flags that had been in the codebase for 8 months.

Upstream offers transparent pricing with a free tier and paid plans:

• Free: $0, includes 1 environment, 10 flags, 10,000 evaluations per month, 7-day audit log, and community support.
• Team: $49 per month for up to 10 seats, includes unlimited flags, 3 environments, 1M evaluations per month, AI Rollout Pilot, Slack/Linear/GitHub integrations, 90-day audit log, and email support.
• Scale: $499 per month with unlimited seats, unlimited environments and evaluations, SSO/SAML, 1-year audit log, 99.99% SLA, dedicated infrastructure option, and priority support.
• Enterprise: Custom pricing with SOC2, HIPAA, on-prem options, custom SLO contracts, dedicated CSM, custom integrations, and 24/7 support.