Captured activity
Connected sources provide the activity available for analysis. Unconnected tools, offline work, and gaps in capture limit the picture.
WorkGraph · Engineering effort allocation
Find the bottlenecks
Find the bottlenecks
no dashboard shows.
WorkGraph uses metadata from connected engineering tools to estimate how each team’s effort is distributed. See work patterns across development, reviews, meetings, and operational work, then identify bottlenecks to investigate.
CalendarPull RequestsTicketsDocsAgent tracesCommits
Trusted by AI-forward enterprises
The problem
The biggest productivity drains don’t show up on any dashboard.
Process overhead creeps up silently. Incidents spill into sprint work. Together, they’re the reason your best engineers feel stuck and the reason quarterly output keeps slipping despite everyone working harder.
WorkGraph makes these patterns visible and actionable.
Effort allocation
Spike detected
Incident response consumed 32% of Platform Team’s time last week.
32%
production incidents, Platform Team
26%
production incidents, all of engineering
24%
product development, Platform Team
Where the week went% of the team capacity%
Production Incidents32%
Product Development24%
Collaboration18%
Code Quality12%
Engineering Design9%
Admin3%
Deployments & Ops2%
Shares estimate captured attention for Platform Team last week, across seven categories. The organization comparison uses the same week and connected sources.
The solution
Engineering work, understood at the right level of detail.
WorkGraph classifies captured engineering activity into seven core categories. Granular enough to spot the real problems. Clear enough to act on them.
Actual observed work, not just surveys.
Beyond the standard view
Your dashboards show output. Productivity blockers show why it’s slipping.
The work that slows your team rarely gets measured. These metrics measure it:
Toil Detection
Surfaces repetitive process work that quietly consumes capacity and is never counted.
Top toil categories
Ticket triage42%
Access requests28%
Manual deploys18%
Backlog grooming12%
Share of the time classified as toil, by the intent behind it.
AI Flow Impact
Shows whether AI tools are improving delivery or just moving the bottleneck downstream.
Engineering week composition
Before AI
After AI
Coding
Review
Meetings
Other
Review grew from 20% to 34% of the week after AI coding tools arrived. Coding shrank 3 points.
Fragmentation Index
Measures real predictors of output quality: context switches, interruption frequency, and uninterrupted blocks.
Interruptions per day
Interruptions per day. The highlighted day had 12, up 10% on the day before.
Developer Intelligence
Shipping
Velocity, quality, cost, delivery confidenceWhat got out the door, and how fast.
What
Sentiment
DX surveys, pulse checks, friction themesHow the team says the week felt.
Why
WorkGraph
How work actually happensWhere the hours went, by category, and which categories were never planned.
How
Each layer answers one question about the same team over the same week, so a change in one can be read against the other two.
Part of 360° Developer Intelligence
WorkGraph completes the picture.
Shipping metrics tell you what. Sentiment tells you why. WorkGraph tells you how and where to act.
Shipping and sentiment are already in Developer Intelligence. WorkGraph adds the third layer, so a slower quarter or a sharper survey comes with the categories that explain it, for the same team and the same weeks.
Explore Developer IntelligenceBuilt for privacy
Team patterns, never individual surveillance.
WorkGraph reports on teams, not people. It reads metadata only, and only from the apps an admin has connected. Message bodies, document contents and code stay where they are.
Always team-level roll-ups · categories are computed per team and per week; there is no per-person view of where the hours went
Metadata only · timestamps, identifiers, participants and sizes. The content of a message, a document or a diff is never collected
Admin-configured apps only · an admin chooses which apps WorkGraph reads. Anything not connected is never touched
Privacy and security
Team-level roll-upsEvery category is computed per team and per week. There is no per-person view of where the hours went.
Always on
Metadata onlyTimestamps, identifiers, participants and sizes. Content is never read, stored or sent to a model.
Enforced
Admin-configured apps onlyWorkGraph reads the apps an admin connects, nothing else. Disconnecting an app stops collection immediately.
Admin controlled
Collected
Timestamps and durations
Ticket, PR and event identifiers
Participants and team membership
Sizes: files touched, attendee count
Never collected
Message bodies
Document contents
Screen recording
Keystrokes
SOC 2 Type IIGDPRSSO & SCIMRole-based accessAudit log
Managers see their own teams, admins see the organization. The same role-based access as the rest of Larridin.
A closer look
How WorkGraph estimates engineering effort
WorkGraph groups captured activity into categories and shows each category’s share of captured attention. This is an estimate of observed work, not a time sheet of every working hour.
Overlap handling
Overlapping activity is reconciled before rollups. Some durations are estimated, so interpret the result as a directional view of captured effort.
Team interpretation
Compare the same teams, categories, and source coverage over time. Investigate a change with the team before attributing it to a bottleneck or an individual.
Questions about WorkGraph
Does WorkGraph track individual working hours?
WorkGraph’s effort-allocation view reports team-level patterns. It does not present an individual time sheet or per-person breakdown of where hours went.
Do the percentages represent the entire working week?
They represent the share of captured attention assigned to each category. Coverage depends on connected sources and observed activity. Offline or unobserved work is not automatically represented.
How is effort allocation different from Engineering Output?
Effort allocation describes how captured activity is distributed. Engineering Output scores eligible merged work. Reading them together helps investigate whether changes in work patterns accompany changes in delivery.
How should I interpret the product examples?
The product views on this page use illustrative data to explain the metrics and workflows. For metric definitions, sample calculations, and assumptions, see our measurement methodology.
Measure what’s changing, in minutes, not months.
AI is reshaping how your engineers spend the week. Start with a discovery call, or try WorkGraph on your own org today.