Ascenda for Organisations
AI made your engineers faster. Learn what it asks of them, and what actually works.
Your delivery dashboards show the output. They don’t show the human × agent work behind it, or how that work landed on the people doing it. Ascenda does, across your engineering cohorts.
The hidden workload
Supervision is the new engineering work.
An AI-assisted hour isn’t one kind of work any more. Some of it is creating. A growing share is directing, verifying and correcting. That supervisory share creeps up quietly, tends to land on your senior engineers, and doesn’t show up in delivery metrics or the quarterly survey.
We measure it from the AI tools themselves, join it to what engineers tell us, and give you a way to test changes against it.
What your organisation sees, at n ≥ 10
AI exposure: how much of the work now runs through agents, and how
Work composition: steering versus supervising, sessions in flight, switching between projects
Interruptions: permission prompts and idle stalls per engineer-day
Time placement: focus blocks, late-day and weekend share
Rework: tool failures, re-prompts and wasted runs
Reported experience: how check-ins spread across strained/clear and drained/activated, and the share of people who read the work as threat rather than challenge. Shown as a direction against the cohort's own baseline, never a score
Two ways to start
A 30-day Report. A 90-day Study.
Each engineer decides whether to join. It’s one question, and it lists exactly what it shares before they answer. Nothing is ticked for them, and they can take any part back later.
30 days
AI Exposure & Work Composition Report
Built from engineers' imported AI-tool history plus live activity. It can run from the command line, without the app. The fastest way to see how AI-assisted work is actually composed across a cohort.
- Exposure and work composition
- Time placement and interruptions
- A snapshot to declare your first change against
90 days
AI Exposure & Work Composition Study
Everything in the Report, live for ninety days, plus check-in self-report and short pulse questionnaires answered in the app. Long enough to score a declared change against a frozen baseline.
- Reported experience in aggregate
- Declared changes and their readouts
- What was measured, and what was not
When the work happened, from the record
Recorded active minutes by weekday and hour, from the tools rather than a survey.
Before the stop
Weeks 1 to 8 · active minutes by weekday and hour
After the stop
Weeks 9 to 12 · same scale
The readout
- After-hours share of active minutes
- 17.5%→8.3%
- After-hours share of active minutes
- Declared measure
- Engineer-days with work after 21:00
- 1 in 6→1 in 40
- Engineer-days with work after 21:00
- Inferred time pressure in the working day
- 44.4→50.8
- Inferred time pressure in the working day
Evening work fell by more than half, and time pressure in the working day rose. The stop moved when the work happened; it didn't make the day lighter.
See it. Change it. Measure what worked.
Declared changes, honest readouts.
Most organisations change something, then look for evidence it worked. A Study has you declare what you expect first, so the readout can’t be fitted to the result.
Declare before you act
Name the change, the measure it should move, the baseline window and the scoring window. The baseline freezes when you declare.
Read it whichever way it goes
The readout is reported in the same place and the same words whether the change worked or not.
Cause, only when designed for
Ascenda names a change as the cause of a move only when the rollout was staggered across cohorts to show it. Otherwise it says what moved, and no more.
Before
Sessions running · a representative day, weeks 1 to 9
9 at once. One line per session in flight at the busiest quarter-hour.
After
Sessions running · a representative day, weeks 10 to 12
4 at once. One line per session in flight at the busiest quarter-hour.
The readout
- Next-day tool failures per active hour
- 1.9→1.1
- Next-day tool failures per active hour
- Declared measure
- Active hours per engineer: the same amount of work
- 13.9→13.6
- Active hours per engineer: the same amount of work
- Sessions running at the peak
- 9→4
- Sessions running at the peak
Fewer agents at once, the same hours, and less rework the next day. The measure moved in the direction the cohort declared.
Coming next: your own delivery figures (the DORA measures) shown beside a readout, as your organisation’s statement rather than Ascenda’s reading.
The wall
Aggregate only. Never an individual.
Every figure covers at least ten people. Below that, it’s withheld. There’s no manager view of any person, at any level of detail. Today every engineer opts in. The rules are published on the Trust page.
Read the published rules →What an organisation never receives
- Any individual, at any level of detail
- A performance, productivity or quality score for a person or a team
- Predictions about a person, such as attrition or flight risk
- Recovery, sleep or any physiological data
- Who joined, who did not, or who left
Start with a cohort of ten or more engineers.
Tell us about your team and the question you want answered. Your engineers can also start with Flow on their own, free, whenever they like.