Common
Project
A real-time canvas that maps client language as it's spoken and lets coaches reframe what the system hears, keeping human judgment at the center.
Real-time NLP.
Human-scale coaching.
This project began with a systems design question: why do master coaches consistently catch meaningful client language that novice coaches miss? During solutions-focused training, I was taught to plot phrases on graphs to track conversational direction, but the cognitive load of note-taking broke presence with clients.
I built a real-time NLP system to extract what mattered and train coaches to listen for it. The breakthrough wasn't better transcription; it was a structured conversational state that could persist between sessions. What started as a coaching aid became foundational infrastructure for AI systems that needed to understand not just what was said, but what it meant in context.
The coach hears everything. The system captures nothing.
During live sessions, coaches had to choose: take notes and miss key moments, or stay present and lose track of conversational patterns. Critical phrases like client contradictions, emotional shifts, and readiness cues would slip past unrecorded. Coaches would miss the moment when a client moved from struggle language (below the line) into solution language, losing the opportunity to amplify that shift and guide them from past strengths toward their preferred future.
Conversational state made visible in real time.
Client phrases extracted verbatim and mapped spatially as they're spoken. Struggles below the line, strengths above. Coaches can see exactly when clients shift from problem language to solution language, and use past strengths (upper left) to fuel movement toward their preferred future (upper right). No notes, no re-reading afterward, just real-time awareness of where the conversation lives on the quadrant.
Not a summary of what happened. A structured view of what was said, legible to the coach and to downstream AI systems.
“I was able to be present with my client for the whole session, rather than pausing to take notes.”
From conversation
to structured insight.
The system processes live Zoom transcripts in real time, extracting meaningful phrases from client speech, classifying them by quadrant, and plotting them spatially. Only phrases that clear the confidence threshold surface on the map, everything else stays in the list layer for coach review. Coaches can move any plotted phrase to a different quadrant by tapping it and selecting a new location. Phrases that are repeated across sessions become more prominent on the map.
Conversation Map · Executive Coaching Session
Four non-negotiables that shaped every decision from confidence thresholds to what the map shows at a glance.
Human authority preserved
The coach sees every classification and can override any of them. The system proposes; the coach decides. This is the design constraint that makes the tool trustworthy in a therapeutic context.
Deterministic over opaque
Every extracted phrase has a visible confidence score and an auditable classification path. No black-box summaries. A coach can always trace why something surfaced and reject it if it doesn't feel right.
Structured memory over transcript storage
The system stores what mattered: extracted phrases, classifications, confidence scores. What gets retained is meaning.
Ambient awareness over dashboard noise
The map is designed to be glanceable during a session. A coach should be able to orient in two seconds and return to the client. Density is the enemy; spatial clarity is the goal.
A transcript records what was said. Conversational state captures what mattered, in what context, with what frequency. Common Project is the layer that converts the former into the latter, the infrastructure that makes downstream AI possible.
Because the coach's attention belongs to the client, not the screen. Ambient awareness is a design goal, not an aesthetic one. Every visual decision, density, color, spatial logic, is optimized for a human who needs to be present, not analytical.