Because a parts catalog shouldn’t just list what you can buy — it should understand what you’re fixing.

The Electronic Parts Catalog (EPC) has been around for decades, but most of them still live in old-fashioned silos — heavy on servers, proprietary data formats, and locked behind subscription walls.
What if we rethought the EPC for the modern era — as a distributed, encrypted, and cooperative app, where suppliers, repairers, and hobbyists all share access to a living, decentralized catalog of vehicle parts?

That’s what I’ve been exploring lately.

The Core Idea

Instead of one central system doing all the work, imagine a distributed EPC where data and logic live closer to the user — panel builders, mechanics, parts suppliers, or DIY hobbyists — each running a lightweight local node of the catalog.

Each node could:

  • Store a local, encrypted copy of the catalog data it cares about
  • Sync updates peer-to-peer via WebRTC or DLT-backed replication
  • Share or restrict data based on role (supplier, customer, repairer)
  • Operate even when disconnected, then sync when back online

In other words: a P2P EPC — fast, resilient, and privately federated.

How It Would Work

The catalog is structured around familiar vehicle hierarchies:

  • Make → Model → Variant/Badge → Year/Month of Build
  • Within each vehicle: Panels, Mechanical Parts, Consumables, Oils, and Accessories

But the magic happens after you identify the vehicle.
Each part links to:

  • Compatible replacements
  • Repair methods
  • Supplier availability (local + remote)
  • Previous purchase history
  • Pricing and warranty metadata

If you’re repairing a specific car, the catalog doesn’t just list parts — it also helps you understand the repair context.
For example:

You look up a 2017 Mazda CX-5 front bumper.
The EPC shows not only part numbers and alternatives, but also:

  • Repair procedures (panel removal, torque specs, safety steps)
  • Associated consumables (clips, primer, sensors)
  • Oils or fluids needed if related mechanical work applies
  • Supplier stock and pricing, across multiple warehouses

It’s not just an index — it’s a repair-aware network.

Finding the Vehicle

Identification can start several ways:

  • VIN or registration lookup via an integrated decoder API
  • Manual selection from make/model/year filters
  • Uploading a previous job or photo record to detect known parts

Once decoded, the EPC automatically filters to compatible components and suppliers.

Every lookup is encrypted, meaning the supplier knows you’re searching for a part type — not your full customer or job context.

The Supplier Network

Each supplier — ACM Parts, independent distributors, OEM resellers — would run a node on the EPC network.
They publish their catalog updates, pricing, and stock levels via secure feeds.
Other nodes subscribe, caching only what’s relevant.

This creates a cooperative network of suppliers, not a single monopoly.
Suppliers control their visibility, pricing, and region. Customers see aggregated availability across multiple providers.

This model takes inspiration from AutoNepo and similar aggregators but pushes it further — by decentralizing the infrastructure, not just the listings.

No single vendor owns the ecosystem — yet everyone benefits from shared visibility.

Privacy & Encryption

Every participant’s data is end-to-end encrypted:

  • Suppliers share only what they choose to (e.g., prices, part numbers, stock)
  • Customers’ search histories and job contexts stay private
  • Repair methods and manuals can be published as encrypted blobs (e.g., IPFS + permission keys)

So the catalog becomes shared but compartmentalized — open in architecture, private in implementation.

Linking Parts to Repair Knowledge

The EPC isn’t just “what part fits.”
It’s how it fits, why it fails, and how it’s replaced.

Linking repair procedures, torque values, and service bulletins creates a knowledge layer — an embedded “repair manual graph” that grows as users contribute.

Think of it as GitHub for repair data: versioned, decentralized, and enriched collaboratively.

Architecture Sketch

  • Frontend: PWA (React + ShadCN UI) — offline-first
  • Storage: IndexedDB (client-side), CouchDB or TimeScale for distributed sync
  • Transport: WebRTC / Kafka-lite for real-time sync events
  • Ledger: Lightweight DLT for verified supplier transactions
  • AI layer: Local LLM for semantic search (“show compatible oil filters for Mazda CX-5 2017”)

This structure allows for local-first compute with federated aggregation, making it both scalable and private.

The Vision

Imagine a world where every mechanic, parts supplier, and car enthusiast taps into a living, encrypted network of automotive knowledge.
No single gatekeeper, no vendor lock-in, no outages taking the catalog down.

Each node contributes to — and benefits from — a collective, verified, distributed EPC.

Fast to load.
Accurate by design.
Private by default.
Shared by intent.

That’s the future of parts catalogs I want to build.

Final Thought

As with most of my projects, this one starts simple — a prototype catalog running peer-to-peer, synced by WebRTC, searchable offline.
From there, the data grows richer, the links deeper, and the network smarter.

It’s an EPC that doesn’t just list parts — it connects the entire repair ecosystem.