SVP – Digital Platforms — Automotive-Chip Business

The mandate

An automotive-chip business is ramping products across external wafer, assembly and test partners. Engineers lose time reconciling lot, die, package, test-program and customer-return identifiers before they can analyse yield. Dashboards disagree because limits and exclusions change without lineage. A planned new SVP – Digital Platforms will create trusted genealogy and analytical services for recovery without claiming that models replace technical disposition.

Approximately 425 employees and material partners depend on the platform across product engineering, quality, operations, applications and suppliers. The SVP owns data architecture, integration, platform products, analytics enablement, access and delivery. Engineering owns yield conclusions and quality owns release; technology preserves reliable evidence and reproducibility.

Genealogy is foundational. Wafer and die identity must survive assembly splits, rework, test insertions and customer shipments. The platform will expose gaps rather than invent links. Supplier data contracts need schema, timing, correction and retention obligations.

Test context matters as much as result. Program version, limit set, equipment, calibration and retest policy determine comparability. The SVP will make these part of lineage and prevent retrospective threshold changes from silently rewriting historical yield.

Analytics will be evaluated for false signal and leakage. Models may prioritise correlation, but engineers must access source populations and understand exclusions. Automotive customer and supplier data require segregation, purpose limits and controlled sharing.

Data timeliness is itself a product requirement. A technically perfect dataset arriving after the lot disposition or customer containment decision has limited value. The SVP will set service levels by use case, design streaming and batch paths appropriately and expose late supplier feeds. Manual emergency data must later reconcile to the governed record.

The platform must support containment. When a suspected failure population emerges, engineers need to identify related wafers, assembly lots, test insertions and shipments without over-including unrelated product. Query results require versioned logic, reviewer approval and preserved evidence because customer and warranty actions may later be audited.

Master data changes deserve controls comparable to code. Product identifiers, bin maps, units, limits and customer mappings can alter analysis across millions of records. The SVP will implement maker-checker approval, impact testing and rollback and will prohibit direct production edits that cannot be reconstructed.

Model lifecycle will include drift and retirement. A classifier trained during early yield may become misleading after process improvement or test-program change. Product engineers will own technical suitability, while the platform records training population, performance, authorised use and review date. Automated scores cannot block or release automotive product on their own.

Operational resilience spans cloud, on-premise and partner connections. Recovery tests must prove that priority genealogy and containment services can be restored within customer decision windows. Backups without immutable source, compatible schema and practised restoration are not resilience. Vendor outages and revoked credentials will be included in exercises.

Cost accountability will distinguish foundational control from optional analytics. Each data product will have an operational owner, consumption evidence and retirement rule, preventing a proliferation of dashboards whose maintenance competes with genealogy and containment during the ramp.

The new role exists because fragmented project teams cannot establish cross-partner authority. The onsite Pune leader will build product-oriented platform teams and serve yield governance.

What you will own

• Establish die-to-customer genealogy across external manufacturing.
• Control test context, schema, lineage and data-quality exceptions.
• Deliver reproducible yield and return analytics for engineers.
• Govern supplier data contracts, access and correction.
• Validate analytical products before operational reliance.
• Protect customer, product and partner confidentiality.
• Prioritise platform work against ramp decisions.
• Build data engineering and product leadership.

The first 12 months

In the first 60 days, trace representative failing and passing units end to end, document missing links and compare disputed dashboards. Set authoritative definitions and interim controls for priority ramps.

By month six, deliver controlled genealogy and test context for priority products, automate exception visibility and establish supplier data agreements. Engineers should reproduce any executive yield metric from governed populations.

At twelve months, achieve 98% complete genealogy on protected lots, reduce analytical preparation time by 70% and cut unresolved data discrepancies by 80%. Platform-supported investigations should shorten root-cause cycles by 30%. No release or customer decision may rely on a model without accessible source evidence and approved technical review.

What the sponsor will measure

• Physical genealogy preserved through outsourced process steps.
• Yield metrics reproducible with test version and exclusions visible.
• Data gaps exposed rather than statistically concealed.
• Engineers spending time on causes instead of reconciliation.
• Partner and customer information appropriately separated.
• Platforms adopted because they answer live ramp questions.

The person

You bring 18–22 years in semiconductor data, manufacturing platforms, product engineering systems or analytics. You have connected foundry, assembly, test and return data and understand automotive traceability. Generic cloud leadership without semiconductor context is insufficient.

Your prior scope should include billions of test records or several outsourced partners. Evidence must include a genealogy break, changed test context and a model whose conclusion you challenged. You can lead technology while preserving engineering and quality authority.

Compensation and terms

Fixed compensation is ₹2.2–3.0 crore plus performance variable linked to genealogy, analytical cycle time, adoption, controls and leadership. This permanent onsite Pune role reports to the Group Chief Executive or sponsor. Timing follows the planned ramp platform.

Confidentiality

The business, products, partners, customer returns, data and models remain confidential. Detail follows fit, conflicts and signed confidentiality. Applicants must not contact suppliers or customers to infer identity.