Live capability demo

Step inside a research run.

Below is one system splitting into a swarm of specialists, each one working a piece of a real research question. Click any ball to step inside that agent and see what it actually did.

This is a ready-made demo. Full autonomous runs take days of compute and real lab access — not something we spin up on page load. What's below is a real, recorded run on a lighter question: do standing desks improve afternoon focus?

Research brief

What Praxis R1 is about to investigate

Before anything runs, Praxis R1 scopes the gap herself and writes it down. This is the brief her swarm will work from.

n = 24 participants 10 workdays Within-subject design Wearable sensors 30 agents · 5 teams

brief_standing_desks_v0.md · Praxis R1

Question

Do standing desks improve afternoon focus?

Gap

Most posture-and-cognition studies run a single lab session. None track real knowledge workers, on real workdays, long enough to see an afternoon slump.

Hypothesis

Standing for part of the afternoon reduces attention lapses without hurting comfort or perceived productivity.

Method

24 participants · 10 workdays · wearable focus-timer + desk-position sensor · paired within-subject comparison.

PRAXIS R1 · RECORDED RUN

STANDING DESKS & FOCUS

Ready when you are

Begin the research run

Praxis R1 will split into a swarm of 30 specialist agents and work the standing-desks question end to end, compressed to about half a minute.

Click a ball to step inside that agent

Agent swarm

30 agents, five teams, working in parallel

0%overall complete

Team timeline00:00:00

Output

What the run produced

Results unlock once the run completes.

Afternoon focus by desk position

Minutes of uninterrupted focus, by time of day and desk position

Focus-lapse rate across the study

Lapses per hour, standing vs. seated, over 10 workdays

manuscript_standing_desks.md · Praxis R1

Standing Desks and Afternoon Focus: A Within-Subject Field Study

Research Brief · OpenScientist · Praxis R1

We tested whether standing for part of the afternoon improves sustained attention among knowledge workers. Twenty-four participants alternated seated and standing conditions across ten workdays while wearable sensors logged desk position and a focus-timer recorded attention lapses.

Standing sessions produced 9% longer uninterrupted focus intervals than matched seated sessions (paired t-test, d = 0.41, 95% CI [0.18, 0.64]), an effect that held after controlling for self-reported caffeine intake. No difference in perceived comfort was found between conditions.

9 citations verifiedreferences.bibConfidence: high

Want to run your own question?

Early access opens in waves. Join the waitlist and invite friends to move up the line.

Join the waitlist →