Yield-Ramp Evidence Board Examiner — Photonics Manufacturing

The mandate

Management reports improving aggregate yield as the company ramps three photonics products, but engineering changes, product mix and test-screen revisions move inside the same trend. Rework and retest recover shipped output while consuming scarce test hours, and customer qualification lots are not comparable with normal production. Directors cannot tell whether the underlying process window is broadening or favourable selection is flattering the curve.

The examiner contributes four days monthly: one reviewing cohort evidence, two challenging factory, product and test teams, and one preparing or attending the committee. Five scheduled meetings are included, with written response within three business days for a material customer-ramp question. Six defined cohorts will be followed from wafer start through packaged test and customer disposition.

The nine-month term ends after the third product reaches its board-approved stability gate. A one-month continuation may be voted only if an approved measurement-system change prevents comparison of the final cohort; renewal requires a restated evidence question and refreshed customer, supplier and investment conflicts. Routine yield monitoring remains management work and cannot prolong the appointment.

The adviser has no line authority and carries no executive responsibility for process changes, lot disposition, production scheduling, customer qualification, test allocation, capital or staffing. Operations and Engineering own the ramp; Quality releases product; the board sets risk appetite. The examiner challenges cohort integrity and causal claims but may not direct experiments, hold lots or promise delivery.

Interests involving photonics producers, foundries, test houses, equipment makers, customers, investors or yield-analytics providers require disclosure. Trading or success-linked exposure to a reviewed ramp is prohibited. The scope excludes engineering review, quality assurance, customer negotiation, capital diligence, vendor selection, individual performance assessment and certification of reported yield.

Why the board wants this voice

Directors receive a single upward yield line even though its population, screen and economics keep changing. They need an experienced ramp operator to challenge comparability and expose rework-dependent output, while leaving process choices, lot release and customer commitments with accountable executives and technical functions.

What you will own

• Press management to define stable cohorts across product revision, route, equipment, material and measurement conditions.
• Test yield claims for mix shift, screening change, retest, rework, exclusion and customer-lot selection effects.
• Challenge apparent progress where test-hour consumption, cycle time or hidden scrap offsets shipped output.
• Examine whether critical defect mechanisms close through evidence rather than disappearing after classification changes.
• Trace six cohorts from wafer release through packaging, test, rework, customer qualification and economic outcome.
• Maintain a committee ledger of confounded trends, open mechanisms, measurement changes and management commitments.
• Probe the ramp with test constraint, supplier-lot shift, screen revision and adverse customer feedback.

Candidate qualifications

• Advised boards or led yield ramps in photonics, semiconductors or another measurement-sensitive advanced process.
• Separated real process learning from mix, screen, rework, retest and selection effects in executive reporting.
• Followed comparable cohorts across wafer, package, test and customer qualification without taking lot authority.
• Challenged defect-closure claims using physical mechanism, measurement confidence, recurrence and full economic consequence.
• Preserved boundaries among board oversight, Engineering experiments, Quality release and customer commitments.
• Maintained independence from foundries, test houses, equipment vendors, customers and investors tied to the ramp.

Non-negotiables

• Available four days monthly for Hsinchu sessions, five committees and six cohort examinations.
• Direct advanced-manufacturing yield governance is required; financial trend review alone is insufficient.
• Will disclose foundry, test-house, equipment, customer, investor and analytics relationships before appointment.
• Will not direct experiments, hold or release lots, allocate test capacity, negotiate customers or certify yield.

1. 49 words maximum. Describe a yield ramp that improved only because cohort or screening definitions changed.
2. 49 words maximum. Which photonics or semiconductor relationships would require disclosure to this board?
3. 49 words maximum. How would you expose output recovered through economically destructive retest?