Taha Adamjee

This portfolio is alive.
AI harvests signals from what I am building and graduates them into stories.

Learn how
←
It's a great time
to be a designer.
A 23-year-old design student in Nottingham who struggles to read his own emails built a device that reads them for him, converts his speech to text, and works in 30 languages. He designed and coded it alone. People without dyslexia started asking for one too. A farmer in Machakos County, Kenya photographs his maize leaves and gets a diagnosis in 20 seconds. He staked his family's season on it. He did not build the tool. The tool earned his trust.

Builders are multiplying. And so are the people who depend on whether what gets built actually holds up when they reach for it.

The work is designing for the relationship between a person and a system. Making ideas usable on the floor, under pressure, at scale. Whether it holds up the moment a real person touches it, under real conditions, with something at stake. Whether it elevates, rather than complicates, the daily lived experience of the people it was meant to serve.

That is what I have held as throughline, over a decade inside Amazon: in operations, learning systems, and emerging technology. There is more to build.

Role Range

  • Sr. UX Designer
  • PM, WHS Learning Innovation
  • Safety Training Project Manager
  • Technical Training Specialist - Tech
  • Program Manager, Associate Training Automation
  • Learning Area Manager, Canada Fulfillment Operations
  • Learning Coordinator, Canada Fulfillment Operations
  • Trainer, Canada Fulfillment Operations
  • Associate, Canada Fulfillment Operations
14 years. Top 0.25% tenure in Canada.
Top 0.7% globally (1.8M employees).
9 roles. 1 company.

Domains

  • Design (throughline)
  • Operations
  • Robotics Safety
  • Learning Operations
  • Fulfillment Launches
  • Learning Systems
  • In-App Training
  • Emerging Technology
  • Mixed Reality

Education

Industry Recognition & Showcase

  • Ops Live 2026
  • Operations innovation showcase, live demo
  • Brandon Hall Gold — Best Advance in XR, 2025
  • Independent global HCM award, analyst-judged
  • Conflux 2025
  • Amazon global design conference, cross-org
  • Ops Live 2025
  • Operations innovation showcase, live demo
  • Brandon Hall Silver — Best Advance in XR, 2024
  • Independent global HCM award, analyst-judged
  • LXD Con Presenter 2022
  • Amazon internal learning design conference
←
Thinking
01

Nature's Playbook

Coexistence Dynamics from the WildResearch

Real partnerships from nature that solve the same problems physical AI faces today.

2025
02

Living Portfolio

AI-Assembled Career NarrativeSpark

What if this portfolio built itself forward as the agent detects milestones?

2026
03

Humorphism

Design Language for Physical AIResearch
2025
Forming
01

Even Realities G2

Smart Glasses AppDevelopment
2025
Made
01

Throughline

This PortfolioLive
2026
←
×

Nature's Playbook

Coexistence Dynamics from the Wild

Two AI agent fleets run autonomously every morning. One surfaces coexistence problems from real operations, policy gaps, and incident data. The other searches evolutionary biology for living partnerships that solve those problems. They feed each other. The library builds itself.

Every dynamic that surfaces passes a 3-gate filter: works across all humans (pre-cultural), both actors benefit (bidirectional), and the vulnerable partner stays safe without needing to understand the system.

I review, elevate, and annotate. The agents learn from feedback, refine coverage, and find the next gap. The system grows daily without prompting.

Autonomous Agent Loop · Daily Generation · Self-Expanding Coverage
The Loop
◎ Scenario Agent (daily, 7am)
│  Reads operations data, policy gaps, incidents
│  Generates 3-5 new coexistence problems
│  Scores each for criticality (severity, probability, gap)
│
├─➔ Nature Research Fleet (daily, 6-8am)
│  Searches biology for partnerships under same conditions
│  Translates behavior into design patterns
│  Maps which problems are now solvable
│
├─➔ Coverage Map updates automatically
│  Gaps revealed → new scenarios generated → new antidotes found
│
└─➔ I review, elevate, annotate
   Agents learn from feedback. Loop continues tomorrow.
01 Sentinel
02 Shared Signal
03 Collapse
04 Boundary
05 Dual Mode
01 · Sentinel Without Consent
The Problem
Nature's Answer
The Design Pattern
A contractor walks onto the floor. No badge, no wearable, no training. The fleet has no data on this person. They are invisible to the system and physically present in the movement zone.
Red-billed oxpeckers ride on rhinos, eating ticks. When a human approaches, the bird shrieks. Rhinos with oxpeckers detect humans 100% of the time at 61 meters. Without them: 23% at 27 meters. The bird attached itself because the rhino offered food. The warning emerged from mutual benefit.
The system attaches awareness automatically. The moment a body appears, it signals the fleet to alter behavior. It signals the stranger through spatial cues they read without training. Protection before comprehension.

How this portfolio works

A living record. AI watches what I build. I decide what stays. Nothing here was placed. It was earned.

The Partnership

I design and lead. AI works alongside me. It watches what I build, learns what gains momentum, and recognizes when something earns a place here. The longer we collaborate, the sharper it gets.

SIGNAL HARVESTER AI agent, scheduled Watches: Slack · Email · Calendar · Knowledge Graph · Files Looks for: recognition, completions, phase shifts │ ▼ EXPLORATIONS living substrate Themes form as signal clusters accumulate Thinking → Forming → Made │ ▼ PORTFOLIO what you see here Stories graduate when evidence reaches critical mass Each strip was earned, not placed

How It Runs

An AI agent runs three times a week alongside my design and leadership work. It scans my messages, emails, calendar, knowledge graph, and local files. It filters for portfolio-grade moments only: external recognition, project completions, and phase shifts. Signals accumulate on the Explorations substrate. When a theme reaches critical mass, it graduates into a story here.

I Decide What Belongs

I approve every graduation. The AI proposes, I decide. I can rearrange, rewrite, and redirect at any point. The intelligence is in the harvesting. The editorial judgment is mine.

This system is constantly evolving. The agent learns what quality means over time. The thresholds adapt. What you see today is one frame of something that keeps moving forward.

←