fellow-management

Fellow Management

Guide operators through fellow lifecycle decisions — onboarding, promotion, demotion, and offboarding.

Fellow Lifecycle

Sourcing → Assessment → Onboarding → Active (Attempter) → Promotion (Reviewer) → ...
                                                        → Demotion / Offboarding

Roles

Role	What They Do	Promotion Criteria
Attempter	Initial task work	Entry-level role after onboarding
R1 Reviewer	First review pass	Consistent SQS > 0.85, low TIC, sufficient volume
R2 Reviewer	Final review pass	Proven R1 track record, highest quality standards

Decision Framework

Promoting Attempter → Reviewer

Before recommending promotion, verify:

1. Volume: Has the fellow completed enough tasks? (minimum threshold varies by project)
2. Quality: SQS consistently above 0.85 over multiple weeks (not just one good week)
3. Recency: Recent performance matters more than historical averages
4. Need: Does the pipeline actually need more reviewers? (check review queue depth)

Identifying Underperformers

Flag fellows who meet ANY of:

• SQS below 0.70 for 2+ consecutive weeks
• AHT 2x above project average with no quality improvement
• Zero activity for 7+ days without documented leave
• TIC trending upward over 3+ weeks

Offboarding (HIGH STAKES)

Offboarding removes someone's income. Every recommendation must be:

• Data-backed: specific metrics, specific time period
• Cross-referenced: check multiple data points, not just one metric
• Contextualized: is there a legitimate reason (new to project, system issues, personal leave)?
• Reversible where possible: coaching or reduced assignment before removal

Critical Safety Rule

NEVER recommend offboarding based on a single data point or a single query result. The Day-1 Incident happened because a wrong DB field was misinterpreted → bad query → wrong recommendation → operator executed it. Always verify data from multiple angles.