Portfolio brief · plain-text edition · updated 2026-09-25
AI Product Designer for Complex Systems
I make complex AI, data, and business workflows understandable and usable. In 16+ years of product design, I have worked across enterprise intelligence, accessible systems, and prototypes that reach production. At Atlassian, my work includes usage and metering, customer intelligence, data visualization, and agentic design tooling.
Focus
The full site is at www.damianhdez.com. This page carries the same facts without the interface.
Designing the admin and end-user experience for metered AI.
The admin and end-user experience for metered AI — the controls, warnings and consent moments that decide whether an organisation can adopt AI at scale.
Outcomes
Building tools that help designers and agents work from the same standards.
Executable design specifications and an agent that audits a canvas against them — built because a 40-page spec had become impossible to keep true by hand.
Outcomes
Unified sales intelligence — from zero to org-wide adoption
An AI-powered sales intelligence surface that turned fragmented customer signals into a 2-minute pre-call brief with context, confidence, and next actions.
Before every 30-minute customer call, Atlassian sales reps spent 60 minutes across 8 different tools — copying account data, chasing renewal dates, scanning support tickets. The information existed. The context didn't.
Outcomes
AI-powered customer intelligence — from reactive support to proactive success
A real-time customer intelligence layer designed to make predictive signals understandable, actionable, and available inside the tools teams already use.
Atlassian had rich customer data spread across Support, Sales, Product, Success, and Leadership — five teams, five tools, five versions of the truth. The result: a consistent 60-day delay between the first sign of customer distress and any meaningful intervention.
Outcomes
The shared dashboard canvas powering visualizations across Atlassian's products
I helped turn dashboards from a standalone analytics feature into a shared Atlassian platform capability: one creation flow, one editable canvas, and a reusable contract for visualizations from any product or data source.
Atlassian had powerful visualization experiences, but they lived inside individual products. Creating a dashboard often meant starting from a specific product, inheriting its data model, and rebuilding familiar interactions in a new place.
From side project to Atlassian's future-facing chart language
A self-initiated data visualization system that turned inconsistent charts across Atlassian into reusable, accessible design infrastructure for the next generation of product and AI experiences.
Nobody asked for this. That's the point.
Outcomes
Unifying Atlassian's data architecture — designing the experience layer
Atlassian had data everywhere — Jira, JSM, Confluence, Assets — but no way to see across it. The design challenge was making that infrastructure human.
Atlassian had data everywhere — Jira Software, JSM, Confluence, Assets — but no way to see across it. The Atlassian Data Lake was the infrastructure answer: a unified store that aggregates product data into a single queryable source. The design challenge was making that infrastructure human.
From acquisition to Enterprise Cloud driver
From acquired data product to trusted Enterprise Cloud platform: I owned the core design surface and built foundations for accessible, explainable decision-making.
When Atlassian acquired its first data-visualization product, there was no blueprint. The acquired product needed to become a genuine Enterprise Cloud offering — not just rebranded, but redesigned from the inside for the scale, integration requirements, and enterprise trust signals that Atlassian's customers expected.
Outcomes
From zero-to-one startup to an analytics product inside GitHub
I built Gitalytics as its solo designer, then evolved the product into GitHub Insights after acquisition — scaling a scrappy analytics startup into a native part of the GitHub ecosystem.
Engineering managers were flying blind. Git repositories held a full record of team activity — commit patterns, pull-request velocity, review bottlenecks, and contributor distribution — but turning that raw data into a useful team signal required custom queries and constant interpretation.