PORTOverview
A leading global client asked for an academic anthropologist chatbot. Instead, we built The Thought Room: a collaborative workspace powered by a multi-agent network of PhD-level AI experts delivering deep cultural insights on demand.
As UX Lead, my challenge was to move beyond the transactional "search box" paradigm. I designed an editorial, research-oriented interaction system that feels like stepping into a physical anthropology studio.
Role & Responsibilities
Process & design pillars
The challenge wasn't to build another chat interface — it was to make AI feel like entering a research studio.
To honour a human anthropological methodology, the UX was anchored on three design principles:
Contextualization
Built a conversational onboarding flow that guides users to reframe surface-level business questions into deep cultural inquiries.
Depth on demand
Created UI handles that let users seamlessly steer the conversation, drilling down into hyper-local nuances or scaling up to macro-patterns.
Reflexivity
Designed interface guardrails that transparently expose AI boundaries, prompting users to critique potential regional or structural biases in real time.
Designing through technical constraints
Design intent must remain intact as the technology scales.
During validation, we hit a performance bottleneck: running image analysis, global data synthesis, and commercial logic simultaneously caused output quality to dip — the V1 Paradox.
To solve this, I kept the V1 UI strictly optimised for anthropological depth — the core value proposition. I then partnered with engineering to map out the V1.5 roadmap, designing UI structures to split the single, overloaded Orchestrator into specialised agent interfaces, ensuring design intent remained intact as the technology scaled.
Impact & validation
Expert approved
Ranked #1 in a blind test against a proprietary AI tool & Perplexity
2/3 external PhD anthropologists noted it "reads the most like an anthropologist."
Strategic shift
From reactive summaries to proactive guidance
Shifted internal workflows from data retrieval toward behavioural intelligence.
Project scale
New internal benchmark for AI-native design
Extended scope into a multi-year capability roadmap.