Overview
Collaborative Sensemaking transforms Condelo's Explore feature from a single-player analysis tool into a multiplayer investigation platform. Multiple users work on the same exploration simultaneously with real-time presence, shared annotations, and AI-assisted research tasks.
This is the difference between analysts working in silos and a team building shared understanding together.
The Problem
- Knowledge work is inherently collaborative, but knowledge management tools are designed for individuals
- Due diligence, incident response, and research projects require teams to synthesize findings from multiple angles simultaneously
- Teams lose context when analysts work in separate tools and manually consolidate findings
- Reasoning processes are invisible: team members see each other's conclusions but not the evidence paths that led there
- Report generation from collaborative analysis is manual and error-prone
How It Works
- Multi-user presence — Extend Explore with real-time cursors and activity indicators showing who's investigating what
- Shared investigation board — A canvas where users pin findings, surfaces, document excerpts, and free-form notes
- Threaded comments — Attach discussions to any pinned item for contextual conversation
- Connection drawing — Users or AI draw typed edges between items: supports, contradicts, depends-on, leads-to
- Research delegation — Assign questions to AI agents or other team members with tracked progress
- Session history — Full audit trail of who contributed what and when
- Report generation — Export a structured report from the board contents with proper attribution
User Story
A due diligence team of 4 analysts evaluates a potential acquisition. They open a shared investigation space. Analyst A deep-dives into financial clusters, bookmarking key findings that appear on the shared board. Analyst B investigates regulatory risks — their annotations are visible in real-time. Analyst C notices a connection between A's financial finding and B's regulatory concern, draws a "depends-on" edge, and adds a comment. Analyst D assigns an AI agent: "Investigate the target company's patent portfolio." All findings accumulate on a shared board that becomes the foundation for their investment memo — generated with one click.
Complexity & Timeline
| Aspect | Detail |
|---|---|
| Complexity | Complex |
| Estimated Build | 6–8 weeks |
| Platform Dependencies | Explore (clusters, conversation, collections), Experiences (report generation), Agents |
| New Infrastructure | WebSocket/Realtime presence, canvas UI, conflict resolution, permission model |
Target Clients
- Personas: Due Diligence Teams, Research Directors, Incident Response Leads, Managing Partners
- Verticals: Private Equity, Consulting, Investigative Journalism, Defense/Intelligence, Law Firms
- Pitch: "Investigate together — see what your team sees, build understanding as a group."
Revenue Potential
Converts Condelo from single-player to multiplayer, creating network effects within organizations. Natural driver for team-based pricing (per-seat within a workspace). Addresses a gap in the market — most knowledge tools are either solo (Notion, Obsidian) or collaborative but shallow (Google Docs, Miro). High retention driver: once a team's investigation workflows live in Condelo, switching costs are significant.
Feature Synergies
- Decision Journal — Team decisions made in shared spaces capture multi-perspective rationale and richer evidence trails
- Expert Fingerprinting — Surface relevant expert reasoning patterns during collaborative investigation
- Embeddable Widgets — Embed investigation board summaries in external project management tools
Risks & Open Questions
- Real-time sync infrastructure (WebSocket or Supabase Realtime) adds operational complexity
- Conflict resolution for concurrent edits on the same board items needs careful UX design
- Permission model (viewer / contributor / admin per investigation) must be intuitive without being rigid
- Canvas-style UI is a significant frontend investment — may need to scope to a simpler list-based collaboration first