Talk

Virtual

Engineering zero failure launches under global scrutiny

How a Staff TPM designed a centralized major launch system governing feature flags, experimentation, rollout sequencing, and per-minute capacity planning to safely deliver 150+ features during live global events with zero failure tolerance.

CEST

Negar Hosseini shares how she transformed a previously decentralized release process into a repeatable major launch practice that powered five global live events serving 260M+ users, achieving 100% execution success with a lean team.

Under intense public scrutiny and zero tolerance for platform instability, she engineered a centralized launch operating model that integrated:
• Defined decision rights and measurable go/no-go criteria
• Unified timelines across 150+ interdependent features
• Governed feature flag management with shared enrollment policies
• Experimentation guardrails and blast-radius budgeting
• Mandatory rollback simulation before rollout approval
• Per-minute traffic modeling and pre-scaling on launch day

Rather than relying on heroics, the system treated feature flags as regulated infrastructure and launch governance as a product capability.

Attendees will learn how to engineer coordination at scale, reduce systemic risk, increase feature concurrency year over year, and enable high-impact launches aligned with live marketing moments without compromising platform stability.

Virtual

Register for PlatformCon 2026