2014 - 2019

Cloudistics Ignite

A UX-first approach to reimagining the private cloud stack

Challenges

I joined Cloudistics as the founding member of the UX team. Product design was situated between two camps.

To my left, executives needed a vision translated to a fully realized private cloud platform. And to my right, I needed to build trust with a pre-formed engineering team lacking front-end expertise and experience working with designers.

Back in the middle, I needed to grow a design team equipped to handle ambiguity in launching a new product and complexity in cloud infrastructure management.

Goal

Our company aimed to provide an innovative turnkey solution for private cloud deployment.

My goal was to deliver a single pane of glass experience for deploying and managing on-premises cloud infrastructures and applications through software-defined composition.

I led the UX team that developed the Cloudistics Ignite management portals, where IT generalists could manage compute, storage, and networking resources and deploy highly scalable workloads.

My Roles

UX Director

I led the Cloudistics UX team that launched and evolved the Cloudistics Ignite platform through experience-first product development.

Product Strategist

I kept the product team on the same page with forward-looking solutions that helped define our product roadmap.

User Researcher

I interviewed internal and external users to improve the user experience.

People Manager

I mentored designers to break down complex problems, articulate design decisions, and assess solutions from detailed and holistic perspectives.

Project Manager

I set stakeholder expectations, outlined team strategy, managed parallel work streams, and guided our cross-functional collaboration.

Design Lead

I set standards of quality through hands-on examples and resolved stakeholder conflicts by pitching simplified concepts and workflows.

Our UX-First Solution

Impact

Leadership

Retained 4 product designers for 5+ years.
Tied for the top manager at the company based on employee surveys.
Awarded the Cloudistics “Founder’s Award” for the “employee with a founder’s mentality, whose tenacity and ingenuity helped make Cloudistics successful.”

“My new manager is not an Ame.
At Cloudistics, we did design reviews better. We did handoffs better. And we had fewer meetings.”

—Former Report

Product

Cloudistics Ignite reduced infrastructure setup time from 4-6 months to just 4 hours.

“Datacenter administrators worried that their jobs would be eliminated by the Cloudistics Ignite solution. We had to change our marketing strategy and approach CIOs instead of their reports.”

—Cloudistics VP of Product

“Right after I was hired, I was supposed to join the VP of Product in a pitch meeting. He couldn’t make it—so I learned how to use Ignite by myself and gave a product demo that saved the client meeting.”

—Cloudistics Sales Engineer

At 50 customers, Lenovo acquired the technology and renamed it the ThinkAgile CP platform.

Fungible acquired Cloudistics in 2019. 🎉

Leadership:

Building a UX Practice

Building a Supportive Routine

To grow a high-performing team, I started with a foundation of psychological safety and a collaborative mindset. In a hectic startup environment, I created a predictable routine to provide space for the team to reflect and exchange feedback on different aspects of our work. We had several regularly scheduled meetings.

Daily Standup – This team-visible pulse point allowed participants to reflect on their daily work and helped me determine who needed unblocking.

Biweekly Critiques – The team met twice weekly to reflect on our progress through design critiques. We reserved some time for social discussions and mid-week updates.

Weekly 1:1 – I provided a dedicated space for each team member to open up and let me know how they were doing and what I could do to help. I provided feedback related to their projects and personal goals.

Yearly Personal Review – My reports and I discussed challenges, attractions, habits, and uncertainties—forces supporting or inhibiting personal growth. This meeting helped my reports set goals and helped me determine how best to assist each person.

Working on the Right Problems

When the product direction was ambiguous, I took the initiative to identify and surface misalignments, opportunities, risks, and tradeoffs as early in the design process as possible. As the UX lead, I focused on developing a shared understanding of the problem and problem space, allowing designers to focus on generating best-fit solutions.

Discussions – I used freeform discussions with the product owner to clarify intent and resolve conflicting product goals. I often presented concept models and process flows to all stakeholders to elicit a deeper exploration of the product strategy.

Product Plan – As a manager, I began writing a product plan for each new feature delivered to engineering. Once executives approved the plan, we expanded it into design documentation that guided the team through the development process.

Job Stories – Job stories were a valuable addition to our problem definitions. Not only did job stories help connect context, motivations, and outcomes in a self-contained unit, but their incompleteness also surfaced gaps in our shared understanding.

Ensuring Smooth Handoffs

Version Control

As a team lead, using version control for our design files allowed me to sleep at night. Engineers appreciated having a “single source of truth” for our production-ready designs.

Review Branches

We sent developers a “design review” branch that served as a space for discussing and updating solutions. Once we—design, the product owner, and engineering—were all on the same page, designers merged their work into the “single source of truth.”

Variation Archives

Our design team took care to save all variations of our work. More than once, a rejected solution served as the springboard for a future product course correction. Twice, the product direction pivoted backward, and our archived work helped us quickly pick up where we left off.

Phase 1:

Discovery

Modeling Concepts and Flows

I created concept models and user flows (right) to facilitate discussions about concepts and relationships within Cloudistics Ignite.

The product combined disaggregated computing resources—servers, storage, networking—each with multiple layers of abstraction and a unique lifecycle.

Diagrams became crucial for building the information architecture from the ground up and engaging in the right conversations with stakeholders and team members.

V1: Focusing on User Goals

In the first version of the Cloudistics Ignite management portal, I prioritized functionality and workflows.

I designed the first version of our UI, which focused on hardware setup, resource allocation, and application deployment—our core functions.

V1 Instance Properties (right) was designed by me.

We learned from this first version and reworked our abstractions and business strategies.

Phase 2:

Design

V2 : Simplifying Flows

In the second version of Cloudistics Ignite, I switched our navigation from the top to the side and added tabs to better scale our growing content.

V2 Instance Properties (right) was designed by me.

Simplifying Disaster Recovery

I designed the disaster recovery feature.

With a paper prototype, I exposed overcomplicated flows and convinced the Founder/CTO to re-architect our disaster recovery workflow.

Initially, the feature called for a branching snapshot tree maintained locally and replicated offsite. Complexity emerged from combining reverting, syncing, and scheduling. The simplified solution involved maintaining a snapshot timeline locally and the full branching snapshot tree at the backup site.

Our CTO recounted this story and emphasized how I saved the company from extensive engineering work when he awarded me the Founder’s Award.

I explored variations in snapshot approaches. This was a precursor to the paper prototype walking through the disaster recovery flow.

The Final Disaster Recovery Flow

Simplifying Oversubscription

I simplified our “oversubscription” concept by questioning repetitive user tasks.

Instead of continually readjusting an oversubscription factor, users would over-allocate resources and review the oversubscription factor. They would get alerted when physical resources run low.

This updated flow fulfilled the business goal of better resource utilization leading to the purchase of additional hardware.

Left: Originally, I used a modal to adjust the oversubscription factor. Right: In the final version, I designed the oversubscription value to be calculated based on allocation inputs.

V3: Improving Efficiency

In the third version of Cloudistics Ignite, we switched from Omnigraffle to Sketch, which enabled us to save time with higher-fidelity (mockup) wireframes.

Our design team was growing. I split project ownership by features and resource types.

I added snapshot functionality to the V3 Instance Properties (right).

I oversaw the end-to-end experience and maintained a high bar for quality of work.

Results

At 50 customers, Lenovo acquired the Ignite technology, and Fungible acquired Cloudistics.

Our experience in UX-first cloud management would help us tackle design challenges launching the Fungible Data Center, a composable cloud solution built on Fungible-invented DPU technology.

Lessons

Stay one step ahead – I take the design further than a comfortable MVP. I have learned that being one step ahead of the current discussion builds trust in my thoroughness. It is a “Goldilocks” approach—by pushing the design further, I can better find a solution that is “just right.”

Surface the gaps – Uncovering knowledge gaps and proposing mitigations help teams avoid wasting effort. Surfacing these gaps early in development yields even greater returns, which is essential in a startup environment.

Make it tangible – Information needs to be tangible to be actionable. Presenting simpler conceptual models and higher fidelity prototypes help product, design, and engineering team members understand the issues and correct course.