EXCEDIFY (Enterprise PLM & AI-Powered APQP Suite)
Led full-stack development for Excedify’s manufacturing quality workflows, linking engineering requirements, bills of materials, risk reviews, and submission approvals.
Trade-offChose a shared multi-tenant Next.js monolith with RBAC over separate per-customer deployments because a single codebase kept compliance behavior and feature rollout consistent across every org, at the cost of tighter tenant isolation than fully separate deployments would give.
Built phase-gate reviews, revision tracking for multi-level bills of materials, and approval workflows for production part submissions. Added AI-assisted document intake and role-based access for supplier collaboration.
Product Screenshots
PLM Suite module grid — Requirements, Part Master & BOM, DFMEA/PFMEA, Gate Management, Project Timing, and Supplier Portals.
PPAP approval workflow, organization/RBAC management, and core engineering applications (CAD, GD&T, Tolerance Stackup, MMC/LMC).
Browser-based 2D CAD tool with GD&T feature control frame annotations.
MMC & LMC calculator for computing maximum/least material condition tolerances.
Tolerance stackup analysis — component input form for nominal dimension, tolerance, and Cpk.
Stackup results with worst-case/RSS tolerance totals and a probability density visualization.
Project-level APQP dashboard tracking phase progress, schedule health, and gate status.
Admin analytics dashboard — tool usage, daily active users, and all-time engagement rankings.
Problem Statement
Traditional automotive and hardware manufacturing teams rely on fragmented, legacy spreadsheets and siloed PLM tools for APQP (Advanced Product Quality Planning). This causes broken traceability between Customer Requirements (VoC), DFMEA/PFMEA risk matrices, Bill of Materials (BOM), and PPAP documentation — leading to critical engineering blindspots, human errors during audits, and weeks of manual administrative overhead.
Engineering Solution & Architecture
APQP 3.0 Core Engine & Phase-Gate Lifecycle: Engineered a 6-stage Stage-Gate review system (Gate 0 through Gate 5) that automatically aggregates dynamic deliverables from active engineering phases, with bidirectional traceability linking VoC, requirements, and engineering drawings via audit-ready RACI matrices.
Generative AI VoC & DFMEA Synthesis Engine: Architected a multimodal AI pipeline leveraging Google Gemini to parse unformatted client VoC documents (PDF, DOCX, XLSX, PPTX) into structured specs, with automated AIAG-VDA 7-Step DFMEA/PFMEA recommendation loops predicting Failure Modes, Root Causes, and Action Priority scores.
Multi-Level BOM Diffing & PPAP Approval Pipeline: Developed hierarchical BOM management (EBOM, MBOM, SBOM) with sub-assembly nesting, revision snapshots, and real-time delta diffing, plus a complete 18-element AIAG PPAP submission engine with dynamic PSW PDF generation and signed QR verification.
Zero-Trust Supplier Portal & Serverless Cloud Infra: Built an isolated external collaboration portal for cross-organizational part sharing and tokenized access control, deployed on AWS via SST v3 with Node.js 24 Lambda streaming execution, presigned S3 asset vaults, and CloudFront edge CDN.