A product design studio combiningstrategy, craft, and AI.
SOLENERY
Connecting homeowners and energy service providers.
We designed a two-sided platform where homeowners explore energy savings, rebates, and qualified providers, while providers connect with potential customers and find new projects.
Services
Product Design · Web Design · Design System
Industry
Home Energy & Clean Technology
Location
Toronto, Canada 🇨🇦
Primary Users
Homeowners · Energy Service Providers
Platform
Web · Mobile
The Challenge
Going solar sounds simple until you try it. Three problems stood between nearly every property owner and going solar:
Incentives are a maze. Programs exist, but they're scattered, inconsistent city to city, and buried in policy language.
Assessments cost money upfront. A traditional energy audit runs $700–800 — paid just to find out if solar makes sense.
Trust is hard to find. Even once the numbers make sense, owners have no simple way to vet an installer.
Before Fill Design
Working with Solenery’s research team, who conducted the interviews directly, we grounded the project in structured discovery on both sides: interviews with homeowners, installers, and contractors, alongside benchmarking against existing energy platforms.
Two hypotheses shaped where we focused:
- Owners had heard of incentives but couldn't access them — the gap was usability, not awareness.
- And because incentives shift seasonally, the product couldn't be a static directory. For providers, the real cost wasn't finding customers — it was finding the right ones without burning ad budget.
From Insights to Features
Research on each side of the marketplace surfaced a distinct set of needs — and each one had to translate into something concrete in the product.
Property Owners
Insight
Feature Direction
Incentives are confusing and fragmented
Unified incentive-matching, personalized to the property (growing focus — see Since Launch)
Contractor pricing feels risky
Transparent, comparable quotes (investigating drop-off — see Since Launch)
Owners don't know where to start
A guided upgrade journey, not a wall of information
ROI is unclear
An instant savings calculator (being simplified — see Since Launch)
Providers
Insight
Feature Direction
Customer acquisition is expensive
Referrals built on qualified leads (now expanded with outreach — see Since Launch)
Weak brand presence
Public profiles showcasing credibility and past work
Inconsistent ROI tracking
A dashboard with clear lead and conversion data
Four personas kept every decision honest: Sarah, a Toronto doctor who wants a straight answer without digging; David, a Vancouver property manager benchmarking multiple buildings; Mike, a Calgary contractor tired of ads that go nowhere; and Anita, an Ottawa marketing manager needing credibility without an agency budget.
Personas
Designing a Two-Sided Platform
Solenery isn't one product — it's two, running in parallel and feeding into each other.
The Owner Flow trades a form for a conversation. It starts with a single promise — "see your energy upgrade plan in under 3 minutes" — then hands off to an AI Advisor that asks one thing at a time, showing its work instead of a blank spinner. What comes back is a personalized assessment — an EnerGuide-style score, a breakdown of where the home loses energy, eligible incentives — rendered as cards inside the same conversation. Only after that value is visible does sign-up come into play, sitting behind a soft lock that still hints at what's underneath. From there, requesting quotes is almost anticlimactic, landing the owner in a dashboard built around one goal: tracking a property's journey from assessment to approved quote.
Property Owner Panel
Solenery AI Advisor
Provider Panel
Provider Panel / Project Page
The Installer Flow opens with a different pitch — business growth, not personal savings — and addresses installers' biggest complaint, unqualified leads, before they even sign up. Installers land in a dashboard built around what needs attention right now; opening a lead surfaces everything a technician needs before a home visit, and a quote upload flows directly into a chat thread with the owner.
The real challenge was never either flow alone — it was keeping an entire matching engine in sync underneath both: incentive eligibility tied to location and season, leads routed by capacity and geography, quotes and messages synced in real time.
The AI Layer
A traditional energy audit requires a technician and an on-site visit, typically costing around $700–800. Solenery’s AI Advisor turns the early assessment into a guided conversation — combining property information with energy consumption patterns to help homeowners understand where their home may be losing energy and what improvements to consider. Rather than presenting a data dump, it translates that information into a prioritized plan with estimated costs and relevant incentives, making a free initial assessment genuinely valuable before moving forward with a professional audit.
AI shaped the design process itself, too — accelerating wireframes, generating supporting assets, and speeding up research synthesis, without replacing design judgment; it simply meant more iterations fit inside the same timeline.
Speed was treated as a design principle in its own right. AI was used wherever it could shorten the distance between an idea and something the client could react to — generating multiple wireframe directions quickly, letting the design team converge internally, and landing on an agreed version within days. That version, not a description of it, is what went into the client meeting — feedback happened on real screens, which made flows concrete enough for Solenery to visualize, react to, and help shape.
Two screens carry that iteration especially visibly.
Getting started. The earliest version asked for a postal code, then walked the owner through a short, classic form of questions about the home. The next version replaced that entirely with an open conversation — an AI Advisor asking for everything, one message at a time, in free-form chat. The version that shipped is a hybrid of both: still a conversation, but questions now arrive as tappable, multiple-choice options instead of free text — a structured wizard living inside a chat.
Understanding the home. The property assessment screen went through the most visible evolution of all — six rounds, from an early grayscale wireframe organized around financing tabs, through a simple dashboard-style property list, to the shipped EnerGuide-style score and energy-loss breakdown. The polish changed least; what actually shifted, round over round, was the hierarchy — which number led, how information was grouped, what the owner needed to see first.
Key Design Decisions
Postal code first. The platform's first ask isn't a name or a form — it's a postal code, framed around a concrete promise ("in under 3 minutes"). That's what makes the "thinking" state right after feel earned instead of fake.
Insight before sign-up. Early wireframes asked for sign-up closer to the quote-request step. The shipped version moves the assessment — score, energy-loss breakdown, incentives — before sign-up is ever mentioned, with the locked screen still hinting at what's underneath.
From wizard to conversation. The first prototypes were step-by-step wizards — functional, but they felt like paperwork. The fix was collecting the same data inside a conversation with an AI Advisor instead, one question at a time, reasoning shown as it happens.
Quotes organized by journey, not by date. The dashboard never shows a flat, chronological list — everything is grouped around where the owner actually is, so "what happens next" is always the answer, not "what happened when."
From abstract to grounded. The first landing page felt too conceptual — people couldn't quite picture what they'd get. The team moved to real photography and copy that names the actual services directly, so the site felt like a service, not a pitch.
Visual Design System
The visual language needed to feel modern and tech-forward, while still carrying the warmth of clean energy rather than the coldness of a typical SaaS dashboard. The system builds on Solenery's existing brand book, extended into landing pages and dashboard panels across desktop, mobile, and tablet.
Typography runs entirely on Inter; color is organized around Sun, Sky, and Forest — three families mirroring the product itself: solar energy, clean air, nature. The one piece of pure brand expression sits inside the AI Advisor: a soft, glowing gradient orb moving through all three tones, giving the AI a warm, recognizable presence instead of a generic spinner.
A full component library — buttons, status tags, data tables, dropdowns — backs both dashboards, with color used purposefully throughout: Sun for financial data, Sky for location, Forest for incentives. Color becomes a quiet second layer of information.
Cross-Platform Considerations
A homeowner isn't at a desk when curiosity strikes — they're on the couch or standing in front of their house. The entire owner flow, from postal code to quote comparison, was designed mobile-first.
Installers move between the truck and the office. Quick lead checks work on mobile; heavier tasks — reviewing property details, preparing quotes, reading analytics — are built around a spacious, desktop-led dashboard.
Two platforms tuned to where their users actually are.
Final Reflections
The real skill with AI in design isn't using it everywhere — it's knowing when to reach for it and when to put it down. The shift from step-by-step wizard to conversational AI Advisor took several real rounds to find, not a single leap.
There's a broader pattern across other AI-driven engagements: teams often default to a chat interface as the primary way users talk to a product's AI, quietly giving up the UX discipline that made the interface usable. Solenery's AI Advisor avoids that trap — the "smart" feeling comes from a flow that avoids repetition and dead ends, not an open-ended chat window standing in for the product.
Just as important was stepping away from AI deliberately, at moments needing nothing but human judgment. AI was the accelerant for exploration; human judgment was the filter for what was worth keeping.
Since Launch — What We're Still Solving
The work didn't stop at launch. A few things have shifted since — and a few are still open questions the team is actively working through.
From landing page to Scout & Outreach. The installer side originally had a dedicated widget and landing page built to attract providers inbound. It didn't perform the way the team hoped, so it was retired — the team replaced it with Scout & Outreach: a feature that finds properties matching an installer's service area and offering, and reaches out to them automatically. Instead of waiting for installers to attract leads, the platform now goes and finds them.
From recommendations to incentives. The product's center of gravity is shifting — early on, the focus was on recommending upgrades; increasingly, the focus is moving toward surfacing and helping owners access financial incentives. That shift is still in progress.
Chasing the quote-request drop-off. One current investigation: fewer owners than expected click through to "Get Detailed Quote." Several rounds of testing — moving where sign-up sits in the flow, rewriting the copy — improved things, but didn't fully explain the drop. Digging into churn data surfaced something simpler: for some owners, the upfront cost estimate is already enough on its own, and a formal quote doesn't feel necessary yet. The current hypothesis: helping them access financial incentives, not a better quote, might be the missing piece.
Removing what confuses. Feedback also flagged that some of the charts shown to owners were adding confusion rather than clarity. That section is being simplified.
Credits
None of this would have moved as fast without Solenery’s team. Their responsiveness made for an unusually smooth collaboration. We worked as an embedded design partner within their product team, collaborating closely through an ongoing retainer rather than a traditional handoff-based engagement. This gave both teams the flexibility to test, learn, and continuously improve the product as it evolved.
Designing AI-driven products has been the defining trend of the last two years — Fill Design Studio has worked on more than five projects where intelligence sat at the core of the product. Most teams who lean on an AI chatbox as the primary interface end up quietly giving up the UX discipline interfaces depend on. That's why talking directly to an AI is treated as a last resort rather than a default — the goal is designing the feeling of intelligence into a product that stays simple, avoids repetition and complex forms, and delivers complete results.
Let’s build what’s next.
Have a product, idea, or challenge you’d like to explore? We’d be happy to hear about it.
A product design studio combiningstrategy, craft, and AI.
work
Services
Lab
ABOUT
Start a conversation
SOLENERY
Connecting homeowners and energy service providers.
We designed a two-sided platform where homeowners explore energy savings, rebates, and qualified providers, while providers connect with potential customers and find new projects.
Services
Product Design · Web Design · Design System
Industry
Home Energy & Clean Technology
Location
Toronto, Canada 🇨🇦
Primary Users
Homeowners · Energy Service Providers
Platform
Web · Mobile
About Solenery
Solenery is a clean energy platform connecting Canadian property owners with clean-energy service providers — a two-sided marketplace where homeowners cut energy costs on one side, and vetted installers find qualified leads on the other. The platform was already live on an early AI-assisted MVP when Fill Design Studio came on to help it scale.
Since launch, it has been featured by outlets including TechBeat Canada, FinSMEs, The AI Insider, Decoder.ca, and Clean Energy Canada.
The Challenge
Going solar sounds simple until you try it. Three problems stood between nearly every property owner and going solar:
Incentives are a maze. Programs exist, but they're scattered, inconsistent city to city, and buried in policy language.
Assessments cost money upfront. A traditional energy audit runs $700–800 — paid just to find out if solar makes sense.
Trust is hard to find. Even once the numbers make sense, owners have no simple way to vet an installer.
Before Fill Design
Before Fill Design
Working with Solenery’s research team, who conducted the interviews directly, we grounded the project in structured discovery on both sides: interviews with homeowners, installers, and contractors, alongside benchmarking against existing energy platforms.
Two hypotheses shaped where we focused:
- Owners had heard of incentives but couldn't access them — the gap was usability, not awareness.
- And because incentives shift seasonally, the product couldn't be a static directory. For providers, the real cost wasn't finding customers — it was finding the right ones without burning ad budget.
From Insights to Features
Research on each side of the marketplace surfaced a distinct set of needs — and each one had to translate into something concrete in the product.
Property Owners
Insight
Feature Direction
Incentives are confusing and fragmented
Unified incentive-matching, personalized to the property (growing focus — see Since Launch)
Contractor pricing feels risky
Transparent, comparable quotes (investigating drop-off — see Since Launch)
Owners don't know where to start
A guided upgrade journey, not a wall of information
ROI is unclear
An instant savings calculator (being simplified — see Since Launch)
Providers
Insight
Feature Direction
Customer acquisition is expensive
Referrals built on qualified leads (now expanded with outreach — see Since Launch)
Weak brand presence
Public profiles showcasing credibility and past work
Inconsistent ROI tracking
A dashboard with clear lead and conversion data
Four personas kept every decision honest: Sarah, a Toronto doctor who wants a straight answer without digging; David, a Vancouver property manager benchmarking multiple buildings; Mike, a Calgary contractor tired of ads that go nowhere; and Anita, an Ottawa marketing manager needing credibility without an agency budget.
Personas
Designing a Two-Sided Platform
Solenery isn't one product — it's two, running in parallel and feeding into each other.
The Owner Flow trades a form for a conversation. It starts with a single promise — "see your energy upgrade plan in under 3 minutes" — then hands off to an AI Advisor that asks one thing at a time, showing its work instead of a blank spinner. What comes back is a personalized assessment — an EnerGuide-style score, a breakdown of where the home loses energy, eligible incentives — rendered as cards inside the same conversation. Only after that value is visible does sign-up come into play, sitting behind a soft lock that still hints at what's underneath. From there, requesting quotes is almost anticlimactic, landing the owner in a dashboard built around one goal: tracking a property's journey from assessment to approved quote.
Property Owner Panel
Solenery AI Advisor
Provider Panel
Provider Panel / Project Page
The Installer Flow opens with a different pitch — business growth, not personal savings — and addresses installers' biggest complaint, unqualified leads, before they even sign up. Installers land in a dashboard built around what needs attention right now; opening a lead surfaces everything a technician needs before a home visit, and a quote upload flows directly into a chat thread with the owner.
The real challenge was never either flow alone — it was keeping an entire matching engine in sync underneath both: incentive eligibility tied to location and season, leads routed by capacity and geography, quotes and messages synced in real time.
The AI Layer
A traditional energy audit requires a technician and an on-site visit, typically costing around $700–800. Solenery’s AI Advisor turns the early assessment into a guided conversation — combining property information with energy consumption patterns to help homeowners understand where their home may be losing energy and what improvements to consider. Rather than presenting a data dump, it translates that information into a prioritized plan with estimated costs and relevant incentives, making a free initial assessment genuinely valuable before moving forward with a professional audit.
AI shaped the design process itself, too — accelerating wireframes, generating supporting assets, and speeding up research synthesis, without replacing design judgment; it simply meant more iterations fit inside the same timeline.
Speed was treated as a design principle in its own right. AI was used wherever it could shorten the distance between an idea and something the client could react to — generating multiple wireframe directions quickly, letting the design team converge internally, and landing on an agreed version within days. That version, not a description of it, is what went into the client meeting — feedback happened on real screens, which made flows concrete enough for Solenery to visualize, react to, and help shape.
Two screens carry that iteration especially visibly.
Getting started. The earliest version asked for a postal code, then walked the owner through a short, classic form of questions about the home. The next version replaced that entirely with an open conversation — an AI Advisor asking for everything, one message at a time, in free-form chat. The version that shipped is a hybrid of both: still a conversation, but questions now arrive as tappable, multiple-choice options instead of free text — a structured wizard living inside a chat.
Understanding the home. The property assessment screen went through the most visible evolution of all — six rounds, from an early grayscale wireframe organized around financing tabs, through a simple dashboard-style property list, to the shipped EnerGuide-style score and energy-loss breakdown. The polish changed least; what actually shifted, round over round, was the hierarchy — which number led, how information was grouped, what the owner needed to see first.
Key Design Decisions
Postal code first. The platform's first ask isn't a name or a form — it's a postal code, framed around a concrete promise ("in under 3 minutes"). That's what makes the "thinking" state right after feel earned instead of fake.
Insight before sign-up. Early wireframes asked for sign-up closer to the quote-request step. The shipped version moves the assessment — score, energy-loss breakdown, incentives — before sign-up is ever mentioned, with the locked screen still hinting at what's underneath.
From wizard to conversation. The first prototypes were step-by-step wizards — functional, but they felt like paperwork. The fix was collecting the same data inside a conversation with an AI Advisor instead, one question at a time, reasoning shown as it happens.
Quotes organized by journey, not by date. The dashboard never shows a flat, chronological list — everything is grouped around where the owner actually is, so "what happens next" is always the answer, not "what happened when."
From abstract to grounded. The first landing page felt too conceptual — people couldn't quite picture what they'd get. The team moved to real photography and copy that names the actual services directly, so the site felt like a service, not a pitch.
Visual Design System
The visual language needed to feel modern and tech-forward, while still carrying the warmth of clean energy rather than the coldness of a typical SaaS dashboard. The system builds on Solenery's existing brand book, extended into landing pages and dashboard panels across desktop, mobile, and tablet.
Typography runs entirely on Inter; color is organized around Sun, Sky, and Forest — three families mirroring the product itself: solar energy, clean air, nature. The one piece of pure brand expression sits inside the AI Advisor: a soft, glowing gradient orb moving through all three tones, giving the AI a warm, recognizable presence instead of a generic spinner.
A full component library — buttons, status tags, data tables, dropdowns — backs both dashboards, with color used purposefully throughout: Sun for financial data, Sky for location, Forest for incentives. Color becomes a quiet second layer of information.
Cross-Platform Considerations
A homeowner isn't at a desk when curiosity strikes — they're on the couch or standing in front of their house. The entire owner flow, from postal code to quote comparison, was designed mobile-first.
Installers move between the truck and the office. Quick lead checks work on mobile; heavier tasks — reviewing property details, preparing quotes, reading analytics — are built around a spacious, desktop-led dashboard.
Two platforms tuned to where their users actually are.
Final Reflections
The real skill with AI in design isn't using it everywhere — it's knowing when to reach for it and when to put it down. The shift from step-by-step wizard to conversational AI Advisor took several real rounds to find, not a single leap.
There's a broader pattern across other AI-driven engagements: teams often default to a chat interface as the primary way users talk to a product's AI, quietly giving up the UX discipline that made the interface usable. Solenery's AI Advisor avoids that trap — the "smart" feeling comes from a flow that avoids repetition and dead ends, not an open-ended chat window standing in for the product.
Just as important was stepping away from AI deliberately, at moments needing nothing but human judgment. AI was the accelerant for exploration; human judgment was the filter for what was worth keeping.
Since Launch — What We're Still Solving
The work didn't stop at launch. A few things have shifted since — and a few are still open questions the team is actively working through.
From landing page to Scout & Outreach. The installer side originally had a dedicated widget and landing page built to attract providers inbound. It didn't perform the way the team hoped, so it was retired — the team replaced it with Scout & Outreach: a feature that finds properties matching an installer's service area and offering, and reaches out to them automatically. Instead of waiting for installers to attract leads, the platform now goes and finds them.
From recommendations to incentives. The product's center of gravity is shifting — early on, the focus was on recommending upgrades; increasingly, the focus is moving toward surfacing and helping owners access financial incentives. That shift is still in progress.
Chasing the quote-request drop-off. One current investigation: fewer owners than expected click through to "Get Detailed Quote." Several rounds of testing — moving where sign-up sits in the flow, rewriting the copy — improved things, but didn't fully explain the drop. Digging into churn data surfaced something simpler: for some owners, the upfront cost estimate is already enough on its own, and a formal quote doesn't feel necessary yet. The current hypothesis: helping them access financial incentives, not a better quote, might be the missing piece.
Removing what confuses. Feedback also flagged that some of the charts shown to owners were adding confusion rather than clarity. That section is being simplified.
Credits
None of this would have moved as fast without Solenery’s team. Their responsiveness made for an unusually smooth collaboration. We worked as an embedded design partner within their product team, collaborating closely through an ongoing retainer rather than a traditional handoff-based engagement. This gave both teams the flexibility to test, learn, and continuously improve the product as it evolved.
Designing AI-driven products has been the defining trend of the last two years — Fill Design Studio has worked on more than five projects where intelligence sat at the core of the product. Most teams who lean on an AI chatbox as the primary interface end up quietly giving up the UX discipline interfaces depend on. That's why talking directly to an AI is treated as a last resort rather than a default — the goal is designing the feeling of intelligence into a product that stays simple, avoids repetition and complex forms, and delivers complete results.
Project Credits
Client
Brantie, Solenery
Product Strategy
Majid Abparvar
Design Lead
Mohadeseh Moslemi
UX Design & Research
Ali Daemi
UX/UI Design
Mo Motabzadeh
Let’s build what’s next.
Have a product, idea, or challenge you’d like to explore? We’d be happy to hear about it.
A product design studio combiningstrategy, craft, and AI.
work
Services
Lab
ABOUT
Start a conversation
SOLENERY
Connecting homeowners and energy service providers.
We designed a two-sided platform where homeowners explore energy savings, rebates, and qualified providers, while providers connect with potential customers and find new projects.
Services
Product Design · Web Design · Design System
Industry
Home Energy & Clean Technology
Location
Toronto, Canada 🇨🇦
Primary Users
Homeowners · Energy Service Providers
Platform
Web · Mobile
About Solenery
Solenery is a clean energy platform connecting Canadian property owners with clean-energy service providers — a two-sided marketplace where homeowners cut energy costs on one side, and vetted installers find qualified leads on the other. The platform was already live on an early AI-assisted MVP when Fill Design Studio came on to help it scale.
Since launch, it has been featured by outlets including TechBeat Canada, FinSMEs, The AI Insider, Decoder.ca, and Clean Energy Canada.
The Challenge
Going solar sounds simple until you try it. Three problems stood between nearly every property owner and going solar:
Incentives are a maze. Programs exist, but they're scattered, inconsistent city to city, and buried in policy language.
Assessments cost money upfront. A traditional energy audit runs $700–800 — paid just to find out if solar makes sense.
Trust is hard to find. Even once the numbers make sense, owners have no simple way to vet an installer.
Before Fill Design
Before Fill Design
Working with Solenery’s research team, who conducted the interviews directly, we grounded the project in structured discovery on both sides: interviews with homeowners, installers, and contractors, alongside benchmarking against existing energy platforms.
Two hypotheses shaped where we focused:
- Owners had heard of incentives but couldn't access them — the gap was usability, not awareness.
- And because incentives shift seasonally, the product couldn't be a static directory. For providers, the real cost wasn't finding customers — it was finding the right ones without burning ad budget.
From Insights to Features
Research on each side of the marketplace surfaced a distinct set of needs — and each one had to translate into something concrete in the product.
Property Owners
Insight
Feature Direction
Incentives are confusing and fragmented
Unified incentive-matching, personalized to the property (growing focus — see Since Launch)
Contractor pricing feels risky
Transparent, comparable quotes (investigating drop-off — see Since Launch)
Owners don't know where to start
A guided upgrade journey, not a wall of information
ROI is unclear
An instant savings calculator (being simplified — see Since Launch)
Providers
Insight
Feature Direction
Customer acquisition is expensive
Referrals built on qualified leads (now expanded with outreach — see Since Launch)
Weak brand presence
Public profiles showcasing credibility and past work
Inconsistent ROI tracking
A dashboard with clear lead and conversion data
Four personas kept every decision honest: Sarah, a Toronto doctor who wants a straight answer without digging; David, a Vancouver property manager benchmarking multiple buildings; Mike, a Calgary contractor tired of ads that go nowhere; and Anita, an Ottawa marketing manager needing credibility without an agency budget.
Personas
Designing a Two-Sided Platform
Solenery isn't one product — it's two, running in parallel and feeding into each other.
The Owner Flow trades a form for a conversation. It starts with a single promise — "see your energy upgrade plan in under 3 minutes" — then hands off to an AI Advisor that asks one thing at a time, showing its work instead of a blank spinner. What comes back is a personalized assessment — an EnerGuide-style score, a breakdown of where the home loses energy, eligible incentives — rendered as cards inside the same conversation. Only after that value is visible does sign-up come into play, sitting behind a soft lock that still hints at what's underneath. From there, requesting quotes is almost anticlimactic, landing the owner in a dashboard built around one goal: tracking a property's journey from assessment to approved quote.
Property Owner Panel
Solenery AI Advisor
Provider Panel
Provider Panel / Project Page
The Installer Flow opens with a different pitch — business growth, not personal savings — and addresses installers' biggest complaint, unqualified leads, before they even sign up. Installers land in a dashboard built around what needs attention right now; opening a lead surfaces everything a technician needs before a home visit, and a quote upload flows directly into a chat thread with the owner.
The real challenge was never either flow alone — it was keeping an entire matching engine in sync underneath both: incentive eligibility tied to location and season, leads routed by capacity and geography, quotes and messages synced in real time.
The AI Layer
A traditional energy audit requires a technician and an on-site visit, typically costing around $700–800. Solenery’s AI Advisor turns the early assessment into a guided conversation — combining property information with energy consumption patterns to help homeowners understand where their home may be losing energy and what improvements to consider. Rather than presenting a data dump, it translates that information into a prioritized plan with estimated costs and relevant incentives, making a free initial assessment genuinely valuable before moving forward with a professional audit.
Getting started. The earliest version asked for a postal code, then walked the owner through a short, classic form of questions about the home. The next version replaced that entirely with an open conversation — an AI Advisor asking for everything, one message at a time, in free-form chat. The version that shipped is a hybrid of both: still a conversation, but questions now arrive as tappable, multiple-choice options instead of free text — a structured wizard living inside a chat.
Understanding the home. The property assessment screen went through the most visible evolution of all — six rounds, from an early grayscale wireframe organized around financing tabs, through a simple dashboard-style property list, to the shipped EnerGuide-style score and energy-loss breakdown. The polish changed least; what actually shifted, round over round, was the hierarchy — which number led, how information was grouped, what the owner needed to see first.
Key Design Decisions
Postal code first. The platform's first ask isn't a name or a form — it's a postal code, framed around a concrete promise ("in under 3 minutes"). That's what makes the "thinking" state right after feel earned instead of fake.
Insight before sign-up. Early wireframes asked for sign-up closer to the quote-request step. The shipped version moves the assessment — score, energy-loss breakdown, incentives — before sign-up is ever mentioned, with the locked screen still hinting at what's underneath.
From wizard to conversation. The first prototypes were step-by-step wizards — functional, but they felt like paperwork. The fix was collecting the same data inside a conversation with an AI Advisor instead, one question at a time, reasoning shown as it happens.
Quotes organized by journey, not by date. The dashboard never shows a flat, chronological list — everything is grouped around where the owner actually is, so "what happens next" is always the answer, not "what happened when."
From abstract to grounded. The first landing page felt too conceptual — people couldn't quite picture what they'd get. The team moved to real photography and copy that names the actual services directly, so the site felt like a service, not a pitch.
Visual Design System
The visual language needed to feel modern and tech-forward, while still carrying the warmth of clean energy rather than the coldness of a typical SaaS dashboard. The system builds on Solenery's existing brand book, extended into landing pages and dashboard panels across desktop, mobile, and tablet.
Typography runs entirely on Inter; color is organized around Sun, Sky, and Forest — three families mirroring the product itself: solar energy, clean air, nature. The one piece of pure brand expression sits inside the AI Advisor: a soft, glowing gradient orb moving through all three tones, giving the AI a warm, recognizable presence instead of a generic spinner.
A full component library — buttons, status tags, data tables, dropdowns — backs both dashboards, with color used purposefully throughout: Sun for financial data, Sky for location, Forest for incentives. Color becomes a quiet second layer of information.
Cross-Platform Considerations
A homeowner isn't at a desk when curiosity strikes — they're on the couch or standing in front of their house. The entire owner flow, from postal code to quote comparison, was designed mobile-first.
Installers move between the truck and the office. Quick lead checks work on mobile; heavier tasks — reviewing property details, preparing quotes, reading analytics — are built around a spacious, desktop-led dashboard.
Two platforms tuned to where their users actually are.
Final Reflections
The real skill with AI in design isn't using it everywhere — it's knowing when to reach for it and when to put it down. The shift from step-by-step wizard to conversational AI Advisor took several real rounds to find, not a single leap.
There's a broader pattern across other AI-driven engagements: teams often default to a chat interface as the primary way users talk to a product's AI, quietly giving up the UX discipline that made the interface usable. Solenery's AI Advisor avoids that trap — the "smart" feeling comes from a flow that avoids repetition and dead ends, not an open-ended chat window standing in for the product.
Just as important was stepping away from AI deliberately, at moments needing nothing but human judgment. AI was the accelerant for exploration; human judgment was the filter for what was worth keeping.
Since Launch — What We're Still Solving
The work didn't stop at launch. A few things have shifted since — and a few are still open questions the team is actively working through.
From landing page to Scout & Outreach. The installer side originally had a dedicated widget and landing page built to attract providers inbound. It didn't perform the way the team hoped, so it was retired — the team replaced it with Scout & Outreach: a feature that finds properties matching an installer's service area and offering, and reaches out to them automatically. Instead of waiting for installers to attract leads, the platform now goes and finds them.
From recommendations to incentives. The product's center of gravity is shifting — early on, the focus was on recommending upgrades; increasingly, the focus is moving toward surfacing and helping owners access financial incentives. That shift is still in progress.
Chasing the quote-request drop-off. One current investigation: fewer owners than expected click through to "Get Detailed Quote." Several rounds of testing — moving where sign-up sits in the flow, rewriting the copy — improved things, but didn't fully explain the drop. Digging into churn data surfaced something simpler: for some owners, the upfront cost estimate is already enough on its own, and a formal quote doesn't feel necessary yet. The current hypothesis: helping them access financial incentives, not a better quote, might be the missing piece.
Removing what confuses. Feedback also flagged that some of the charts shown to owners were adding confusion rather than clarity. That section is being simplified.
Credits
None of this would have moved as fast without Solenery’s team. Their responsiveness made for an unusually smooth collaboration. We worked as an embedded design partner within their product team, collaborating closely through an ongoing retainer rather than a traditional handoff-based engagement. This gave both teams the flexibility to test, learn, and continuously improve the product as it evolved.
Designing AI-driven products has been the defining trend of the last two years — Fill Design Studio has worked on more than five projects where intelligence sat at the core of the product. Most teams who lean on an AI chatbox as the primary interface end up quietly giving up the UX discipline interfaces depend on. That's why talking directly to an AI is treated as a last resort rather than a default — the goal is designing the feeling of intelligence into a product that stays simple, avoids repetition and complex forms, and delivers complete results.
Project Credits
Client
Brantie, Solenery
Product Strategy
Majid Abparvar
Design Lead
Mohadeseh Moslemi
UX Design & Research
Ali Daemi
UX/UI Design
Mo Motabzadeh