Pyxos
Pyxos is an AI-driven PII data compliance and DSR automation platform.
I was the first designer brought on board to get the MVP off the ground and work toward the Alpha phase. Our goal with the MVP was to validate initial concepts and present something to customers for feedback. After that, we focused on designing interfaces for the more complex aspects of the product. Once we finalized the wireframes, we began testing user flows with customers and creating high-fidelity designs to hand off to engineering. We tackled the projects in phases, collaborating closely with our R&D teams as the technology evolved.
​
Before I joined, the team had spent several months conducting market research into the problem space. This groundwork gave us a clear vision of what we wanted to solve and the order in which we wanted to address each issue in the first iterations of the product.
Product Planning and MVP
Team
1 - Head of Design (manager and IC) Me
1 - Senior Product Designer
1 - Senior Researcher
2 - PMs
A lot - Engineering & Data Science
Stakeholders
Executive team
Compliance
Legal
Data Science
R&D
Requirements
-
Onboarding
-
Compliance specific customizable, multi-agent chat
-
Chat history
-
Business settings
-
Document Context Center
-
Create entry points for coming soon features and gauge user interest
Getting Started
I kicked this off by gathering and organizing all of the research that had already been done and spending several in-person working sessions with engineering to understand the technical workflows and functionality of the product. We would not initially build this into the MVP, but begin planning for the direction of the Alpha in the MVP. I spent some time understanding the research and organizing thoughts and user flows for both phases.
Approach
Simplify complexity as much as possible.
Solution
We iterated quickly and shipped the MVP within 3 months of kicking off the project.
​
Impact
Customers found the compliance guidance valuable and accurate in the chat. We were able to gain feedback on future features which guided how we prioritized the alpha. Below is a screenshot of the general design and recording of live MVP product. The MVP was intentionally not pixel perfect.
​
Safehouse Connectors and DSR (deletion request) Dashboard
Requirements
-
Safehouse detail (NDA, designs not shown)
-
Onboarding connectors and schema to Safehouse (NDA, designs not shown)
-
Connector status dashboard
-
DSR (PII deletion request) dashboard
-
Deletion request detail
-
Filing exceptions
Approach
We approached this by first focusing on the Safehouse screens and the DSR dashboard. The connector onboarding piece would be done manually behind the scenes to test PII deletion functionality. Once that work was done we ran several rounds of user testing to validate UX, terminology and functionality. With very good signals we moved forward on nailing down the PII data connector flows and also ran several rounds of user testing. I was most nervous about this as it was extremely technical and there was a lot of ambiguity in exactly what needed. The goal was to not overcomplicate the UX and keep it simple. To mine (and most of the teams) surprise, user testing went remarkably well and we felt confident to move forward building this experience. Research conducted by our Sr. Researcher.
Note: I have not included detailed views of the designs for the safehouse configuration and connectors due to NDA.
Privacy Policy Analysis Prototype ​​
This project was one that rated very high outside of the core functionality of the product. We worked together with product, data science and engineering to align on technological constraints and how we build the first iteration. After aligning on flow and UX, we delivered hifi designs to engineering for implementation. Hifi designs completed by our Sr PD. I acted as manager for this project.
DSR (deletion request) Dashboard Prototype​​
The core of Pyxos' product is automating compliance and management of customer PII. A customer must first onboard their PII system connectors and setup their safehouse through our setup wizard (not shown NDA). Once their systems are set up and appropriate deletion rules defined, the dashboard is where an admin can monitor all deletion requests and handle any issues that might arise in the process. The first round of user testing came back extremely positive. We made some adjustments, mainly removing the risk meters and ratings. We learned most companies did not want that until they could customize and define their specific risk parameters. I completed all of the design work for this project.
Jarvis Conceptual Navigation Prototype ​​
