Skip to content
Evan Baker.
← Back to portfolio

Real-World Delivery

Centralized B2B Support Escalations

One escalation path for 10+ partner travel platforms, run on a single severity model and SLAs that every partner played by.

20% faster bug resolution · 15% higher SLA compliance · ~99% availability · No added headcount

Role
Technical program manager
Context
Travel technology platform
Year
2025–26

[ Problem ]

Every partner had its own escalation path and its own idea of what counted as urgent. Partners, Commercial, and Engineering didn't agree on severity, so engineers often worked on whoever pushed hardest instead of the issue with the biggest business impact. Nobody had a reliable view of what was open or who owned it. Prioritization was a negotiation.

[ Approach ]

  • Centralized every B2B production issue through my team, giving 10+ partner platforms one intake and triage workflow, worked by my team of four support engineers.
  • Introduced one severity model and set of SLA guidelines for every partner. A few teams pushed back at first and came around.
  • Launched a Slack bot that auto-triaged support requests, created Jira tickets, and suggested troubleshooting steps while tickets waited for an owner.
  • Designed KPI reports, built by my team from Grafana data, that went to Commercial, Product, and Engineering leadership every two weeks.
  • Wrote post-launch handoff playbooks for three partner launches. A partner moved from its launch manager to my team only once escalation paths, stakeholder contacts, certificate renewals, dashboards, contractual SLAs, infrastructure, API connections, and deployment processes were covered. I also worked with the NOC to tune monitoring and alerting.
  • As we grew from 2 strategic partners to 10+, RCA requests piled up, so I built a Glean agent that drafts partner-specific RCAs from the incident record, the Slack war room, and the post-mortem doc. An RCA went from 1 to 2 hours to about 10 minutes, review included.

[ Outcome ]

  • Average bug resolution time dropped 20% and SLA compliance rose 15% across the partner portfolio.
  • Held about 99% availability with no added headcount.
  • First contact resolution went up 15% once the triage bot was live.
  • Our most strategic partner cut its recurring bug meetings down to once a week.
  • Partner updates got clearer, and leadership got KPI reports on a regular schedule.
  • Prioritization stopped being a negotiation.

[ Tools & skills ]

Incident & Escalation ManagementSeverity FrameworksSLA ManagementTechnical Program ManagementStakeholder ManagementKPI ReportingPlaybook & Runbook DocumentationWorkflow AutomationAI AgentsGrafanaJiraSlackGlean