Export-grade package passport for Indian OSATs that compresses chip qualification and lot traceability into one workflow.

1. Commercial output at Sanand means qualification pain starts with real units and real customer approvals now, not after years of policy discussion.
2. The path from 200 million chips today toward 5 billion annually and a second G2 facility means process debt from manual qualification workflows will compound quickly.
3. Early exports to Japan and Malaysia force export-grade traceability and customer evidence from the first years of operation, before local teams have deep packaging playbooks.
4. Sources explicitly frame OSAT as India's strategic foothold and call out qualification, workforce, and advanced-packaging bottlenecks, which makes execution software the next missing layer.
5. One back-end stack serving automotive, telecom, industrial, EV, and IoT programs multiplies customer-specific approval work enough to support a sharp workflow wedge.

Catalyst. CG Semi has already started commercial output, shipped initial export batches, and begun scaling toward multi-billion capacity inside an OSAT-first national push, which makes manual qualification workflows an immediate bottleneck rather than a future one.

Package Passport OS sits above existing tester, quality, and line records to create a live evidence graph for each chip program. It ingests lot history, test results, deviation records, operator signoffs, and customer-specific qualification requirements, then auto-builds the approval pack a Japanese or Malaysian buyer actually wants to review. When a yield excursion or returned lot appears, the product traces which process window, equipment, or shift changed and opens a controlled requalification workflow instead of sending teams back into email archaeology. The first deployment is one export program at one plant, measured on time-to-approval and number of rejected documentation loops. Over time, the same graph becomes the system of record for cross-line transfer, second-facility ramp, and multi-site customer audits.

What's different. MES vendors record what happened on the line, and QMS tools store documents after the fact. Package Passport OS wins by mapping those records into customer-specific qualification flows, export evidence, and requalification decisions for each chip program. Its moat compounds as it learns which evidence packages, deviation patterns, and approval sequences actually move new programs through overseas customers faster.

Jobs to be done

flowchart LR
  Buyer[Quality head at Indian OSAT] --> Pain[Manual export qualification and traceability]
  Pain --> Product[Package Passport OS]
  Product --> Outcome[Faster chip-program approvals and fewer rejected lots]

Executive takeaways

• The wedge is real because India’s first greenfield OSAT and ATMP plants are shipping or qualifying output before their internal qualification memory is mature.
• The product should land as a read-only evidence and approval layer above tester, MES, and QMS systems rather than as a rip-and-replace MES.
• India-first is strategically attractive but commercially narrow; venture-scale outcomes require expansion into adjacent packaging hubs and broader change-control workflows.
• Competition is strong in adjacent MES, quality, yield, and low-code workflow layers, but few incumbents own the customer-facing qualification pack itself.
• Adoption risk is mostly about data integrity and customer acceptance of digitally assembled evidence, not about whether qualification work exists.

Market definition

Software that turns OSAT and ATMP line, test, and quality records into customer-ready qualification and requalification evidence for semiconductor packaging programs.

Customer and buyer

Primary users are the quality, NPI, and test-engineering teams at ramping assembly and test sites. The economic buyer is usually the plant GM, operations head, or quality leader because approval speed affects shipments, utilization, and customer trust.

Buying triggers

• Pilot-line output or a new package program needs customer approval before mass production can scale. [3][12][20]
• A second line or fast capacity ramp overwhelms a lean quality team with PPAP, traceability, and deviation-evidence work. [5][6][7][8][37]
• Automotive or industrial buyers demand documented traceability, change control, and qualification packs after a process or package change. [13][14][18][20][21][22]

Willingness to pay

Willingness to pay is credible because these plants are already deploying enterprise manufacturing and quality systems alongside multibillion-dollar capacity ramps; shaving approval delay or requalification rework protects scarce line time and export credibility. [5][6][7][17][23][26][28]

Category dynamics

Growth signal >80% CoWoS capacity CAGR (2022-2027) in the tightest advanced-packaging bottleneck

Validation signals

• CG Semi’s pilot line is explicitly framed around customer qualification before large-scale production.
• Micron expects Sanand to move from tens of millions of chips in 2026 to hundreds of millions in 2027, proving the pace of backend scale-up.
• Tata, Kaynes, and Mitsui/AOI-linked announcements show more than one Indian group is building packaging/test capacity and cross-border support.
• Automotive quality tooling remains formal and documentation-heavy, which supports a specialized workflow layer.
• NIST is elevating semiconductor traceability and provenance as a current industry problem, not a distant one.

Regulatory & technical constraints

• Automotive-facing programs still require standardized PPAP and APQP documentation and approval gates.
• Traceability has to survive packaging, marking, and downstream issue investigation at lot or unit level.
• Yield excursions or product/process changes can trigger requalification and customer communication workflows.
• Advanced packaging ramps amplify data-integration burden because package, substrate, memory, and test contexts all matter.

The field is crowded in adjacent layers—MES, quality, yield analytics, and configurable frontline apps—but the open space remains the buyer-facing approval record that assembles plant evidence into customer-specific qualification and requalification decisions.

Why incumbents do not win by default

• Semiconductor MES suites. MES vendors capture execution and genealogy well, but they do not win by default because customer-facing qualification logic and export evidence packaging are not their core workflow.
• Yield and test-data platforms. Yield platforms are strong at data capture and excursion analysis, but they stop short of owning the approval pack, reviewer routing, and exception-closure loop.
• Flexible frontline workflow platforms. Low-code manufacturing tools can build traceability apps quickly, but semiconductor-specific qualification semantics and approval-history logic still require deeper domain modeling.
• In-house quality stacks. Plants can stitch together PPAP portals, spreadsheets, and document libraries themselves, but that leaves genealogy, deviations, and customer approvals fragmented across tools.

Package Passport should launch as a read-only qualification and requalification layer for Indian OSAT and ATMP plants that have already started export shipments or are about to qualify them, not as another MES. The first customer is a Gujarat- or Assam-based plant quality leader juggling 2-5 queued chip-program approvals, export customers in Japan or Malaysia, and fragmented evidence across tester logs, MES records, QA spreadsheets, and PDF packs. The product's job is narrow: assemble lot genealogy, test results, deviation history, and operator signoffs into a customer-specific qualification pack plus a requalification trail fast enough to keep scarce line capacity shipping. Research supports the pain and the timing because India is standing up multiple greenfield assembly-test sites, automotive and industrial programs still require formal PPAP-style documentation, and packaging is becoming more strategic as advanced packaging demand rises. The near-term market is real but concentrated, with research sizing TAM at roughly $120.0M, SAM at $6.8M, and a reachable three-year SOM of $1.8M across four plants, so this is a credible workflow wedge but not a stand-alone venture outcome unless it expands into second-facility transfer, supplier genealogy, and adjacent Asian packaging hubs. The best early proof is not revenue alone; it is whether one design-partner plant can cut qualification-pack turnaround and rejected documentation loops without asking customers to abandon PDF-based approval habits. The biggest gaps in the research are actual qualification volume per plant, which system currently owns tester and deviation records, and whether export buyers will trust linked digital evidence behind a PDF pack. That uncertainty makes a pre-seed plan appropriate: fund 18 months to win 2-3 design partners, prove data completeness and buyer acceptance, and decide whether the company is becoming a category system of record or a useful but bounded workflow feature.

Problem

• New OSAT plants can ship commercial output before they have repeatable qualification memory, leaving quality teams to manually stitch together lot genealogy, test data, deviation history, and approvals for each export or new-device program.
• When a yield excursion, package change, or customer rejection occurs, the requalification loop falls back to email, spreadsheets, and vendor logs, slowing approvals and consuming margin-critical line time.

Solution

• Ingest tester, MES, QA, and operator records in read-only mode to build a live package passport for one chip program, then auto-assemble the customer-specific qualification pack and drill-down traceability evidence the plant already has to send.
• Turn excursions and process changes into a controlled requalification workflow with versioned evidence, disposition history, and legacy-ready PDF outputs so the product fits current buyer habits before asking customers to adopt digital-first review.

Why we win

• We sell the approval record, not the execution stack. MES, yield, and low-code tools capture data, but none of the named competitors is centered on the export qualification pack, reviewer routing, and exception-closure loop itself.
• A library of pack templates, approver behaviors, and linked genealogy-to-outcome history compounds into a proprietary approval dataset that plants and incumbents do not have in one place today.
• India's greenfield OSAT wave lets us land before local process habits harden, while the same workflow extends naturally to second-facility transfer and adjacent Japan/Malaysia-linked packaging hubs if the first plant works.

Milestones

flowchart LR
  Wedge[India first-export OSAT wedge] --> MVP[Read-only package passport MVP]
  MVP --> Proof[Faster approvals and fewer rejected packs]
  Proof --> Expansion[Multi-site transfer, change control, adjacent Asian hubs]

Founding team

Experiment roadmap

Risk assessment

What must be true

• Target plants run at least 6-8 new qualification or requalification events per quarter with enough pain that manual pack assembly delays shipments or line utilization.
• One live export program can be integrated read-only across tester, MES, and QA systems in under 8 weeks without forcing a rip-and-replace MES project.
• Japanese, Malaysian, or automotive buyers will accept passport-generated packs when the output preserves their current PDF approval format and adds drill-down evidence behind it.
• A pilot can cut qualification-pack turnaround or rejected documentation loops by at least 30% within one quarter.
• At least 2 of the first 5 target plants convert paid pilots into annual contracts and create credible reference accounts for adjacent hub expansion.

Open diligence questions

• What is the actual quarterly volume of qualification packs, requalifications, and documentation rework at 4-6 named target plants?
• Which system of record owns tester data, deviation history, and customer document packs today, and how incomplete are those records?
• Will the first buyer fund this from quality or operations budget, or does it require a separate IT or MES approval cycle?
• What network boundary—cloud, customer-managed VPC, on-prem, or air-gapped—does a target plant require before approving a pilot?
• Why does this remain a standalone workflow instead of a feature inside Critical Manufacturing, camLine, Siemens, PDF Solutions, or Tulip?
• What specific expansion path turns a $6.8M SAM wedge into a larger company before incumbents bundle it?

Model sanity

• Revenue engine. Base revenue comes from converting 3 India design-partner plants into annual contracts by Q4Y2, then adding 1 adjacent-hub paid plant in Q1Y3 and expanding each plant toward roughly $450K ARR.
• Must go right. Read-only integrations and PDF-accepted output must standardize enough that 7 ending FTE can support 4 paying plants while gross margin still climbs into the low-70s.
• Model breaks if. If adjacent-hub expansion slips and plant ARR stalls closer to $400K, the downside case keeps EBITDA deeply negative and pushes cash toward the low-$400K range.
• Next-round proof. The seed story is 3 annual plant contracts plus 1 adjacent-hub paid proof of concept with a validated customer-managed VPC or on-prem deployment pattern and late-Y2 burn below about $35K per month.

Scenarios

Sensitivity

flowchart LR
  TargetPlants[Target India and adjacent-hub plants] --> PaidPilots[Paid design-partner pilots]
  PaidPilots --> AnnualContracts[Annual plant contracts]
  AnnualContracts --> ExpansionModules[Transfer and change-control modules]
  ExpansionModules --> Revenue[Revenue]
  Revenue --> GrossProfit[Gross profit]
  GrossProfit --> Cash[Cash and runway]

Flags: Y1 revenue includes onboarding-heavy design-partner pilots, so early revenue should not be read as fully recurring ARR. · The base case still tops out at 4 paying plants by year 3, so a larger venture outcome requires adjacent-hub and second-facility expansion to work beyond India. · Gross margin only reaches the target band if read-only connectors, PDF templates, and deployment-boundary reviews become repeatable; customer-specific VPC or on-prem requests can keep services drag high. · The model uses EBITDA as a proxy for cash and ignores working-capital timing, which is reasonable at this scale but can understate cash needs if large deployments require upfront compliance or hosting work.

• Data integrity gap. If line, tester, or QA records are incomplete, customers may not trust the package passport enough to base approvals on it. Mitigation: Start read-only, score evidence completeness, and integrate the highest trust data sources before automating any approval workflow.
• Beachhead concentration. India's first wave of OSAT ramps may produce only a small number of near-term buyers, which could slow revenue before the category broadens. Mitigation: Use Indian design partners as proof points, then expand to other new OSAT ramps and multi-site transfer workflows where the same qualification pain appears.
• Incumbent system absorption. MES or quality-suite vendors could add basic qualification document management and make the category look like a feature. Mitigation: Own the cross-system customer-approval workflow, the approval-history graph, and export-specific requalification logic that plant systems do not capture well today.