top of page

Accelerating decarbonization with AI driven intelligence

Writer: Google X - Tapestry Energy
Google X - Tapestry Energy
Mar 30
4 min read

Updated: Apr 7

Product Design Lead

3 years 9 months

User Research, User Interviews, Competitive Analysis, Task Flows, Journey Mapping, Design Ops, UX, Mockups, Prototypes, Moderated Usability Testing, Field Research, Visual Design, Design Systems, Branding/Identity, Wireframes




CHALLENGE

Grid operators managing 20+ year infrastructure decisions were working across fragmented, disconnected tools with no unified view of the thousands of data points needed to model scenarios and interconnect renewables at scale.



STRATEGY

As a design team of one, I was a key contributor in conceptualizing and launching two core applications — the Grid Planning Tool and GridAware — establishing the research, design, and product foundations for a platform now used by nation-state-level utilities across three continents.



IMPACT

Unlocked over $2M+ in partnership funding, grew the design org from 1 to 4, and delivered CAPEX savings of hundreds of millions for energy partners including AES (US), Vector (New Zealand), and CEN (Chile).



CHALLENGE

Grid Planning Tool


The paradox of scale


Grid planning is slow, fragmented, and high-stakes. Planners modeling 20+ year infrastructure decisions had no unified view of the data they needed — line congestion, asset health, interconnection timelines all lived in separate systems. Scenario comparison was manual. For decisions that would shape infrastructure across long time horizons, the gap between the tool and the task was significant.



STRATEGY

Research-led, system-built, validated fast


My process started with understanding grid planners and their workflows. As the design and research lead, I planned and executed over 10+ research studies across the engagement. The foundational study was a deep-dive into the user journey of energy systems planners — capturing goals, daily activities, and the pain points that slowed them down most. Those interviews fed directly into product requirements, task flows, and the CUJs (critical user journeys) that anchored every wireframe that followed.




The tool itself had to serve two distinct user groups: distribution and transmission planners. Transmission planning required modeling scenarios stretching 20+ years into the future. Distribution demanded a different kind of granularity. I designed separate experiences for each, consolidating thousands of data points into a single unified view where planners could select, compare, and investigate scenarios and mitigation projects with confidence.





With the user foundation in place, I ran an agile branding and identity workshop with senior stakeholders to establish a distinct visual identity for the tool by modifying design system tokens like color system and typography within the Google Material Design component library. This created two advantages: it preserved engineering efficiency and differentiated the product from a generic Google solution.





With the interim design system in place wireframes quickly transitioned to mockups. For validation and alignment, I created interactive prototypes before MVP launch, with each round of feedback incorporated before the next. Staying aligned with a mixed EPAM and Google engineering team — in an environment where requirements shifted as the technology evolved — required active communication across every channel.






After launch, I planned and executed moderated user testing to take the big leap from testing to adoption. The sessions served as both a learning guide and feedback session giving the team valuable insight into rapid optimization.





CHALLENGE

GridAware


A grid too large to inspect


To preserve the reliability of the distribution grid, utility companies routinely inspect physical assets — poles, transformers, power lines — to catch issues before they become outages. But the inspection process itself was the problem.


Inspections are highly manual and time-consuming — taking 2+ years for a full inspection cycle — which means utilities are often only able to inspect a fraction of their assets each year.


Inspectors worked in the field, often in rural areas with no connectivity. Parts of their toolset were analogue and required work arounds for the conditions they operated in. And without a reliable, up-to-date model of the distribution grid, utilities were making reliability decisions on incomplete information.



STRATEGY

From geo-mapping tool to field-ready intelligence platform


At the start of my engagement, GridAware was already in use by several US utilities. Described by users as a cross between Google Maps and Facebook, it allowed planners to geolocate assets and view lifecycle information from installation to decommissioning. The foundation was there — my goal was to evolve it into something meaningfully smarter.





I established a feedback intake process so critical issues were triaged quickly rather than lost in a backlog, and conducted ongoing usability testing to surface what users needed most. The first major research-driven feature was offline mode — field interviews revealed inspectors routinely lost connectivity in rural areas, blocking access to the app entirely. I worked with product owners and engineering leads to design and launch an MVP that let users cache asset data before heading into the field.


Through a strategic partnership with Vector, New Zealand's grid operator, I planned and conducted field research directly with inspectors and managers. Those sessions became the foundation for personas and CUJs that fed directly into product requirements — ensuring what we built reflected how the work actually happened.


The most transformational output was an ML-powered inspection workflow. By automating image collection and running machine learning analysis to identify infrastructure defects, GridAware could flag issues at a scale no manual process could match. I designed the human-in-the-loop interface for validating ML annotations — preserving human judgment where it mattered most, while dramatically accelerating coverage.




Having an up-to-date model for the power distribution grid, grants utility companies the capabilities they need to ensure grid reliability for their communities.



RESULTS


$2M+ Unlocked


$2M+ in funding unlocked via product and UX based contractual milestones with nation-state level partnerships like CEN of Chile


Patent Holder


Recognized within a small group of inventors included for pending patent:


Integrated scenario modeling for electric grid decisions … [IDF-51829]


Testimonial


“You were a one-person design team, spearheading multiple products and supported UX research for a long time and it was truly remarkable. I’ve deeply appreciated working with you…”


 
 
bottom of page