Pulsar shapes every message around each participant's real-world data: their Fitbit steps, their weight from the scale that morning, their timezone, their response patterns, their place in the study. You define the rules once. Their data decides the rest.
The screen your team opens in the morning, live from your first enrollment. Today's messages, steps, and weights each sit against this study's own daily average. Delivery, reply, and device sync rates carry eight-week sparklines, so a declining pattern shows up weeks before it shows up in your data. Whatever needs a human is already waiting in the alerts panel.
Everything you need to run sophisticated message-driven research studies.
Their week and day in the study, their engagement measured against this study's own rates, the full message thread in both directions, and exactly which device connections are live.
Scheduled message topics and momentary prompts live together, with a timeline across the top showing which rules overlap in any given week. Every card carries its audience, its active weeks, and how often it fires.
Pulsar detects problems before they become crises. Root-cause diagnosis tells you exactly why something failed - landline, carrier block, or missing device data - with inline actions to resolve.
Standard and AI-generated reports ready when you need them. Export history with timestamps for IRB compliance. Download CSV data for your own analysis.
Design sophisticated EMA and survey-sampling protocols with a guided visual wizard. No coding required.
Select how participants receive prompts: fixed schedule for evenly-spaced times, random schedule for unpredictable sampling, or threshold-based to fire when a health metric falls below a target.
Then walk through scheduling, message content, response windows, and skip conditions, all in a guided wizard. Review a plain-English summary before going live.
Message pools are a simple yet super effective way of defining the messages participants receive. Choose sequential delivery (in order) or random (a different message each time). A pool can reset automatically after all messages have been exhausted, or remain in an empty state so that no more messages will be sent from that pool.
Add, edit, or reorder messages anytime. Each prompt can include survey-style questions that participants respond to directly.
Signal-contingent, interval-contingent, and event-contingent sampling. Random prompts within defined windows or fixed schedules.
Trigger surveys based on heart rate, step counts, or sleep data from connected wearables. Threshold-based prompting built in.
Respect individual sleep schedules. No prompts during sleep hours. Automatic adjustment per participant.
Native mobile notifications with configurable reminders. Response windows prevent stale data collection.
Define when to skip or delay prompts based on recent responses, sensor data, or inbound keywords. Delay or reset modes.
Real-time dashboards show response rates, completion times, and participant engagement. Spot declining patterns early.
Upload the allocation list your statistician produced, stratified however your design calls for. It is held as data, so it can be exported and audited later.
We will generate one for you too, and tell you plainly that generating it here suits unblinded, low-risk work only.
Turn off arm reveal and assignment disappears from participant pages, from Insights answers, and from every export automatically. Per-arm counts still work, because a count shows the balance of the trial, not anybody\'s assignment.
Managing the list and reading it are separate permissions, so a coordinator can maintain a list they are not allowed to open.
PHQ-9, PHQ-2, GAD-7, GAD-2, AUDIT, AUDIT-C, PCL-5, PC-PTSD-5, PHQ-15, and WHO-5 ship ready to attach. So do a HIPAA authorization and an informed consent template with placeholders for your IRB's approved language.
Start blank, or start from a validated instrument and add your own questions around it. Just don't reword the scored items, so your results stay comparable to published norms.
Every instrument's setup in one place: whether it gates enrollment, how it scores, which instrument it chains to next, and the text that goes out with the link.
Score each answer, set a threshold, and chain to a follow-up instrument that only fires when the threshold clears. A brief screener leads to the long one, but only for the people who qualify.
Publish any instrument to a public URL with no login required, with a QR code to match. That's how a flyer works.
Already built it in Qualtrics? Link it, and each participant gets their own personalized link on your schedule.
Describe the protocol you want and review what it drafts. Ask a question about your own data and get an answer. You approve everything before a participant sees it.
Use our internal Anthropic engine to personalize messages based on each participant's journey. Your data doesn't go anywhere; our AI engine is built-in so your study remains secure. You review everything before anything is sent out, so you can let the computer do the grunt work while supervising and ensuring the best outcome for your study.
Ask questions in plain English. Get instant answers across messages, Fitbit, weights, and nutrition data. Save queries as live stats, export to CSV, or bookmark for quick access.
Mark any template for AI generation. Personalized content is written at the moment of delivery using real participant data.
Pulsar connects to the tools your study already uses.
Schedule a chat! No pressure to commit. We're happy to answer questions and to chat about your study compatibility with our platform.