SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 248,000 - 279,000 / annual
Discord is seeking an Engineering Manager to lead the Safety Processing team, which builds and operates the systems that review reported content, enforce policies, and handle user appeals at scale. You will manage nine engineers organized into three squads: Review (automated content moderation systems), Enforce (policy enforcement and appeals infrastructure), and Tooling (analyst-facing systems). Each squad has an experienced tech lead, and you'll report to the Director of Engineering for Safety.
The team has achieved significant automation gains—most safety reviews are now decided by LLM pipelines rather than humans, reducing cost per review by over 90% while maintaining or exceeding human-level precision and recall. The remaining work focuses on automating ambiguous, high-stakes decisions and handling hundreds of thousands of appeals fairly and quickly.
Key responsibilities include: leading three squads toward near-complete automation of review decisions while maintaining quality; rebuilding the appeals system to provide consistent paths for all eligible appeals and tooling that measures decision quality; owning investigator tooling that reduces complex investigations from hours to minutes; ensuring enforcement traceability so users and staff can understand how enforcement decisions were reached; managing regulatory and child-safety compliance work alongside product roadmap; setting engineering standards with your tech leads; and hiring and developing the team.
You'll work closely with content review operations, machine learning, product, policy, and legal teams. The role carries regulatory attention, legal deadlines, and serious incidents as part of normal operations. The charter is larger than current headcount, and the team is expected to grow.
Required: 5+ years as a software engineer with strong backend and distributed systems experience, plus 2+ years managing engineers. You must have shipped LLM systems in production where output drives business decisions (not suggestions), with opinions on evaluation, ground-truth data, quality regressions, cost, and latency. You're technical enough to review designs, follow hard incidents, and respectfully challenge approaches without taking work from senior engineers. You have experience working with operations or human-in-the-loop teams, understand quality assurance and reviewer calibration, and build tools that respect their users. You remain steady under escalation and communicate effectively across engineering, ML, data science, policy, legal, and product. You're an empathetic recruiter and coach.
Bonus: passion for Discord and gaming communities, prior trust and safety/fraud/abuse engineering at consumer platforms, experience building internal tools for operations teams, or experience consolidating legacy systems or managing outsourced operations.
