Skip to main content

Team Analytics

Team Analytics provides dashboards for team admins and owners to monitor their team's usage of ProdE. The dashboard is organized into four tabs — Overview, Codebase Chat, MCP, and Dashboard Events — each surfacing engagement trends, user activity, and feature adoption through interactive charts and tables.

All tabs support date range filtering and an option to exclude internal users from the data.


Overview

The Overview tab gives you a high-level snapshot of team engagement across all ProdE channels.

MetricTypeWhat It Tells You
Active UsersKPITotal unique users who interacted with ProdE in the selected period across any channel.
Active Users by ChannelTableBreaks down active users by channel — Codebase Chat, Dashboard, and MCP — so you can see which surfaces your team uses most.
Active Users by DateTime-series chartDaily trend of unique active users, with lines for each channel. Helps you spot adoption trends and usage dips over time.

Codebase Chat

The Codebase Chat tab drills into how your team uses ProdE's chat interface for code-related tasks.

KPIs

MetricWhat It Tells You
Active UsersNumber of unique users who used Codebase Chat in the selected period.
New ChatsTotal chat sessions initiated. Each new conversation counts as one chat.
MessagesTotal user messages sent across all chats.

Charts

ChartTypeWhat It Tells You
Active Users by DateTime-seriesDaily active user count with breakdown by source (Web, Slack, Jira). Shows where engagement is coming from over time.
New Chats by DateTime-seriesDaily count of new chat sessions. Useful for tracking adoption and identifying spikes in usage.
Messages by DateTime-seriesDaily message volume. Indicates how deeply your team engages with the chat — more messages per chat signals deeper exploration.
Chats by SourcePie chartProportional split of chats by where they were initiated — Web, Slack, or Jira.
Messages by SourcePie chartProportional split of messages by source. Helps you understand which platform drives the most conversation volume.

Tables

TableColumnsWhat It Tells You
Active Users by SourceSource, CountHow many unique users are active on each source (Web, Slack, Jira).
Tool Use CountsTool Name, CountWhich AI tools are invoked most frequently within chats (e.g., code search, file read). Shows which ProdE capabilities your team relies on.
Chats by UserUser Name, Email, CountPer-user breakdown of chat sessions. Identifies your most active users and those who may need onboarding support.
Messages by UserUser Name, Email, CountPer-user message counts. A high message count relative to chats indicates deeper, more complex conversations.

Chat Objective Analysis

ProdE uses an LLM to automatically categorize each chat conversation by its primary intent. This helps you understand what your developers are using the chat for — not just how much they use it.

Objectives are grouped into categories:

CategoryObjectivesWhat They Mean
DiscoveryFind API Endpoint, Find Implementation, Understand Architecture, Understand Data Model, Find ConfigurationDevelopers exploring and navigating the codebase.
PlanningPlan Feature, Plan Migration, Design ArchitectureDevelopers using ProdE to plan upcoming work.
DevelopmentWrite Code Snippet, Write Code Detailed, Debug Issue, Root Cause AnalysisActive coding and debugging tasks.
ReviewCode Review, Security Review, Performance AnalysisQuality assurance and review activities.
DocumentationCreate Documentation, Understand Documentation, OnboardingDocumentation-related usage, including new developer onboarding.
TestingTesting Strategy, Write Test CasesTest planning and generation.
ExternalExternal Integration, Web Search InfoLooking up external APIs or searching for information.
DecisionCompare Approaches, BrainstormingEvaluating options and ideation.

You can drill down into any objective to view the individual chats categorized under it, along with summaries of each conversation.


MCP

The MCP tab tracks how your team uses ProdE through IDE integrations (e.g., Claude Code, Cursor, Windsurf). MCP (Model Context Protocol) tool calls represent direct AI-assisted interactions from within the developer's editor.

KPIs

MetricWhat It Tells You
Active UsersNumber of unique users who triggered MCP tool calls in the selected period.
Tool CallsTotal number of MCP tool calls made. Each invocation of a ProdE tool from the IDE counts as one call.

Charts

ChartTypeWhat It Tells You
Active Users by DateTime-seriesDaily count of unique users making MCP calls. Tracks IDE integration adoption over time.
Tool Calls by DateTime-seriesDaily volume of tool calls. Spikes may correlate with active development sprints or new feature rollouts.

Tables

TableColumnsWhat It Tells You
Tool Call BreakdownTool Name, CountWhich specific MCP tools are used most frequently (e.g., ask_specific_codebase, get_high_level_context). Reveals which ProdE capabilities are most valuable in the IDE workflow.
Tool Calls by UserUser Name, Email, CountPer-user tool call counts. Identifies power users and helps gauge individual adoption of IDE integrations.

Dashboard Events

The Dashboard Events tab tracks how your team interacts with the ProdE web dashboard — including document views and feature engagement.

KPIs

MetricWhat It Tells You
Active UsersNumber of unique users who accessed the ProdE dashboard in the selected period.
Document ViewsTotal number of documents viewed across all sources.

Charts

ChartTypeWhat It Tells You
Document Views by SourcePie chartSplit of views between Docs Page (browsing documentation directly) and Chat (documents surfaced during a chat conversation). A high "Chat" proportion means developers are discovering docs organically through AI interactions.
Document Views by DateTime-seriesDaily document view volume. Helps track documentation engagement trends over time.