DraftThis is a tour of every module in ProductIntel. Each surface shown here is in production. Polished individual concept pages are being added one by one.
The Platform, Designed
The other concept pages show what the platform does in motion. This page shows what it contains. Every module is grouped the way it appears in the sidebar of the running product. Each one gets a small designed snapshot so you can see the actual surface, not a feature bullet.
The point of consolidating all of this onto one platform is not feature count. It is that every artifact — story, doc, case, request, agent run — lives in the same graph. The modules are different lenses on the same data.
The four surfaces a PM and team open every day. Backlog, knowledge, support, and the morning triage.
AI triage tells you what needs attention before you ask.
/dashboard
This morning · 3 attention items
Backlog, sprints, board, smart assignment, AI triage.
/work
Sprint 47 · board
Backlog
14
In Progress
3
Review
2
Done
6
Documentation organized by product. Powers the Context Engine.
/knowledge
Recent docs · 4 of 47
Customer support hub with case → story handoff and chat origination.
/cases
Active cases · 12
Five surfaces for composing, supervising, and measuring an agent workforce alongside humans.
AI-narrated weekly briefing across all agent activity.
/workforce
“This week · 47 agent runs across 6 teams. Token spend trending down 12% week-over-week. Two agents earned new badges. The Account Health Analyzer flagged a customer at risk of churn.”
Trading-card view of every agent — specialty, tier, badges, stats.
/workforce/roster
Roster · 4 of 21
★★★
Discovery Agent
Senior · 1,247 runs
★★
Product Owner
Mid · 893 runs
★★★
Observer
Senior · 2,104 runs
★
Persona Agent
Junior · 218 runs
Compose teams with strategy, supervisor styles, and guardrails.
/workforce/teams
6 teams · grouped by capability
Real-time activity stream of running agents.
/workforce/monitor
Live activity
Per-agent performance + Training Arena for A/B testing team configs.
/workforce/performance
This week · top 3 by quality score
Three surfaces that watch your backlog, your industry, and your team — surfacing patterns on their own schedule. See the Monday brief concept page for the centerpiece view.
Unified Monday brief converging Discovery, Strategy, and Behaviors.
/intelligence
“Today · 3 of 47 candidate signals surfaced. 2 Discovery + 1 Strategy + 0 Behaviors above threshold.”
Topic-driven competitive and market research with AI-generated briefs.
/strategy
Active topics · 8
Subscribe to industries — Reddit, HN, RSS — filtered by product context.
/intelligence
Subscribed feeds · 6
Learned heuristics from how your team operates over time.
/skill-wallet
“Active patterns · 14. Most recent: "Maya leads visualization stories — 10 of 12 sprints." Confidence 91%.”
Three surfaces for managing accounts, fielding requests, and giving customers a chat interface that connects back into your knowledge.
Lightweight CRM with health scoring, contacts, and deal pipeline.
/accounts
At-risk accounts · 4
Public roadmap with voting, feature requests, NPS surveys.
/admin/feature-requests
Top requests · this month
External-facing RAG chatbot for end customers, built on your knowledge.
/admin/consumer-chat
“Acme Insurance support chat · 23 sessions today. Average 2.4 turns to resolution. 4 escalated to human · 1 spawned a case → work unit.”
Three surfaces for keeping the AI workforce honest, observable, and safe.
Side-by-side model comparison on golden test cases.
/admin/calibration
Recent calibration · spec-enrichment task
Trace every LLM call with retrieval sources, prompts, and reasoning.
/admin/inference
“Trace TR-7218 · spec-enrichment · 12 retrieval sources · 4,231 input tokens · 892 output · $0.018. Reasoning chain captured. Open to inspect prompt composition by source.”
8 automated scanners with auto-fix pipeline and dashboard.
/admin/security
Security · last 24h scan
Everything is configurable from the admin UI — agent prompts, model selection, module toggles, work mode. No code changes required for any of it.
Agent registry, model configuration, module toggles, onboarding.
/admin
Configuration surfaces · 9
The Whole Toolbox
The other concept pages show what one cycle looks like. This page shows the rest of the toolbox the cycle runs inside. The Discovery Agent on the Intelligence Hub reads from the same graph that Cases writes to. The Roster pages reflect the same agents that Monitor watches. The Inference Inspector traces calls that Knowledge surfaces. Every module is a different lens on the same underlying data.
That is the bet from Chapter 02 of the homepage made tangible. Federated stacks cannot do this. A single instance with one graph can.