Skip to content
All projects
SystemsFull StackConcept / Academic

EWaste Hub

A team-designed platform connecting e-waste producers, collectors, and recyclers, with AI-assisted batch analysis and compliance support.

This is a concept / academic project — it has not been deployed as production software.

SYSTEMS
Product DesignSystem ArchitecturePRD Writing

Project Overview

EWaste Hub is a team project focused on the coordination problem in e-waste management: producers, informal collectors, and licensed recyclers operate with very little shared visibility into volume, location, or compliance status. The team produced a full product requirements document covering the connecting platform, an AI batch-analysis feature, and NEMA compliance documentation support.

The Problem

E-waste in this context moves through fragmented, largely informal channels. Producers don't have a reliable way to reach compliant collectors, collectors lack tools to batch and report what they've gathered, and recyclers have limited visibility into inbound volume — all of which makes regulatory compliance (with bodies like NEMA) harder to track and enforce.

Target Users

E-waste producers (businesses and institutions), informal and formal collectors, and licensed recycling facilities.

Goals

  • Design a platform that connects all three actors in the e-waste chain with role-appropriate tools
  • Specify an AI-assisted batch analysis feature to help collectors and recyclers characterize incoming e-waste volume
  • Build in NEMA compliance documentation as a first-class feature rather than an afterthought
  • Include a Disposal Cycle Reminders feature to nudge producers toward timely, compliant disposal

Challenges

  • Reconciling the needs of formal (licensed, compliance-driven) and informal (mobile, cash-driven) collectors within a single product
  • Defining what 'AI batch analysis' should actually do without overstating what's feasible for a team-scoped PRD
  • Mapping NEMA compliance requirements accurately into concrete product features rather than vague 'compliance support' language

Proposed Solution

The PRD specifies a role-based platform: producers list disposal needs and receive reminders on a defined cycle, collectors accept and batch pickups with AI-assisted volume/category estimation from photos, and recyclers receive structured intake records that map directly onto NEMA documentation requirements.

System & Technical Architecture

Producer portal → Collector matching & batch intake (with AI-assisted classification) → Recycler intake & compliance reporting, with Disposal Cycle Reminders running as a scheduled notification layer across the producer side.

Producer Portal
Collector Matching & Batch Intake
AI Batch Classification
Recycler Intake & NEMA Reporting

Key Features

  • Producer-to-collector matching by location and waste category
  • AI-assisted batch analysis for volume and category estimation
  • NEMA-aligned compliance documentation generation
  • Scheduled Disposal Cycle Reminders for producers

Technical Decisions

Treated compliance documentation as a core feature, not a report bolted on afterward

Regulatory documentation is often what determines whether an e-waste actor can operate legally at all — building it into the core data model made the compliance output a natural byproduct of normal platform use rather than a separate manual process.

Testing Strategy

As a PRD-stage project, validation was done through stakeholder analysis and structured requirements review with the team and mentor rather than software testing.

Security Considerations

Not yet applicable this project is a specified concept and has not been implemented as running software.

Results & Outcomes

A complete product requirements document exists, covering user roles, the AI batch-analysis concept, and NEMA-aligned compliance features. This project has not been built as functioning software. [This is a concept / academic PRD project, not a deployed platform.]

Lessons Learned

  • Designing for both formal and informal actors in the same value chain forces genuinely different feature sets, not just different UI skins
  • Regulatory requirements are easiest to satisfy when they're designed into the data model from the start, not retrofitted

Future Improvements

  • Prototype the collector batch-intake flow as a working application
  • Validate the AI batch-classification concept against real e-waste photos

Want to talk through how this was built, or a similar problem you’re facing?

Get in touch