UX CASE STUDY · CLIMATE TECH
Aire Labs x Purpose Design Co.
UX Research & AI-Augmented Systems Synthesis for Climate Tech
In June 2024, I joined Purpose Design Co. for a three-week research sprint with Aire Labs, exploring how climate tech project developers make decisions under complexity, and what it takes for AI-enabled tools to earn trust in that process.
The work focused on surfacing where expert assumptions diverged from real user needs, using qualitative research and systems-level synthesis to inform product direction.
TEAM
Cross-functional research team
(Founder & 4 researchers)
MY ROLE
UX Researcher
TIMELINE
3 Weeks
METHODS
Qualitative Synthesis
AI-assisted clustering
Insight Framing
DELIVERABLES
Affinity Mapping
Empathy & Personas
Presentation
The Challenge
Aire Labs was building an AI-enabled software platform to support climate-positive project development, leveraging predictive models and geospatial visualizations to help teams deploy, finance, and scale complex systems.
The initial research intent focused on understanding how Project Developers might use advanced tooling to:
Evaluate Scenarios
-
Evaluate scenarios
Forecast key variables
Manage Financial Risk
-
Manage financial risk
Advocate for decisions with
evidence and clarity
The goal was to inform an early product vision for an AI-driven platform embedded within real project development workflows.
My Role & Contribution
As a UX Researcher at Purpose Design Co., I focused on qualitative synthesis and sensemaking across a fast-moving discovery sprint.

When I joined the project, the research goals and interview structure had already been established. My responsibility centered on synthesizing interview data and translating complex qualitative input into strategic insight.
This included:
-
Structured note-taking across stakeholder and user interviews
-
Affinity-based synthesis to surface systemic patterns
-
Identifying signals that informed shifts in research focus
-
Translating findings into actionable insights for cross-functional teams

This work required integrating into an active research stream and rapidly building shared context across teams.
💡RESEARCH PIVOT
From Product Capabilities to System Constraints
As interviews progressed, a clear pattern emerged.

While conversations were framed around tools, technology, and AI-enabled capabilities, participants consistently redirected toward deeper, systemic challenges that shaped their day-to-day work.

Rather than centering on specific features or visualizations, Project Developers emphasized:

-
Non-linear, ad-hoc workflows
-
Relationship-heavy coordination
-
Manual processes driven by trust gaps
-
Risk aversion tied to opaque automation

This reoriented the entire research focus.

Instead of validating or refining a predefined solution space, the work moved toward understanding the underlying systems and constraints any successful tool would need to respect.
This pivot allowed the research to surface root-level insights that reframed both the problem space and future opportunity areas. 

From this synthesis, a clearer design focus emerged.
"How might we better support climate tech project developers as they navigate complex, relationship-driven workflows - while helping them feel more confident in the decisions they make every day?"
AI-Augmented Research & Synthesis
Given the volume and density of qualitative input, I incorporated AI as a pattern-surfacing assistant.
SYNTHESIS APPROACH
RAW CAPTURE
400+ qualitative notes captured through live transcription and review
SYNTHESIS TRANSITION
AI-assisted clustering used to surface early thematic hypotheses
AI-ASSISTED CLUSTERING
Human-led validation to refine, challenge, and contextualize patterns
FINAL SYNTHESIS
Final synthesis distilled into five system-level insights

This approach reduced cognitive overhead while preserving judgment, allowing deeper attention to meaning, edge cases, and risk.
What Emerged: System-Level Insights
01
Project development workflows are inherently non-linear
Participants described their work as iterative, ad-hoc, and shaped by evolving inputs, stakeholders, and constraints. Rigid tools often fail to reflect this reality.
02
Manual processes persist due to trust gaps
Despite awareness of AI and automation, developers default to manual workflows when outputs lack transparency or explainability.
03
Relationship management is core infrastructure
Much of the work depends on negotiation, credibility, and informal coordination, which are critical dynamics that existing tools rarely support.
04
Information quality outweighs information volume
The challenge is not access to data, but confidence in its reliability and interpretation.
05
Standardization is desired when adaptable
Participants want shared frameworks that flex by role, context, and project stage rather than enforcing one-size-fits-all processes.
This research suggests that opportunity in climate-tech tooling lies not in aggressive automation, but in augmented intelligence.

Successful systems in this space:
-
Respect non-linear workflows
-
Make reasoning and outputs transparent
-
Support relationship-heavy decision environments
-
Reduce cognitive burden without obscuring nuance
AI has meaningful potential here if it’s designed to support human judgment.
Reflection
This project marked my first formal UX research engagement within climate tech, a domain marked by complexity and uncertainty.

Rather than positioning myself as a subject-matter expert, I focused on pattern recognition and epistemic restraint: holding space for ambiguity while allowing insights to emerge through evidence.

That approach continues to shape how I work today, especially in systems where trust, risk, and human judgment are central.
How I Use AI in Research
AI is most useful to me as a sensemaking partner.
In this project, I used AI to reduce cognitive load during synthesis by surfacing early pattern hypotheses across a large volume of qualitative data. These hypotheses were treated as starting points, not conclusions.

I rely on human judgment to validate, challenge, and contextualize what AI surfaces, especially in domains where trust, risk, and nuance are central.
In practice, this means AI helps me see faster, while I remain accountable for what ultimately gets framed as insight.
AI supports my work by:
-
Accelerating early clustering
-
Reducing synthesis overhead
-
Highlighting potential pattern blind spots
Human judgment remains essential for:
-
Context preservation
-
Ethical interpretation
-
Final decision-making