Ad-driven volunteers chat with an AI pre-screening bot that walks each inclusion and exclusion criterion, scores eligibility against the protocol, and books qualified candidates straight into coordinator calendars — with their full answers attached — while polite disqualifications keep the database warm for future trials.
Typical results
24/7
Screening coverage
-70%
Coordinator phone time
12%
Screening no-show rate
Figures reflect typical results for this kind of automation.
The challenge
Every ad campaign flooded the site with volunteers, but coordinators could only phone-screen during office hours. Most callbacks reached voicemail, interested candidates cooled off within days, and coordinators spent twenty hours a week asking the same eligibility questions — mostly to people the protocol excluded in the first two minutes.
Before and after
Qualified candidates per month
3×Coordinator phone-screening
−70%Screening-visit no-show rate
−60%The automation
We built an AI pre-screening bot connected to the recruitment ads: volunteers land in a chat that walks through the trial's inclusion and exclusion criteria in plain language. OpenAI scores each conversation against the protocol, qualified candidates pick a slot directly in the coordinator's calendar, and the full transcript with structured answers attaches to the booking. Near-misses and exclusions get a courteous message and file into a database segmented for future studies.
How it works
- 01
Ad click to chat
A volunteer taps the recruitment ad and lands straight in a conversation — no form, no callback wait.
- 02
Criteria walk-through
The bot asks the trial's inclusion and exclusion questions in plain language, one at a time.
- 03
Eligibility scored
OpenAI scores the answers against the protocol and flags anything borderline for human review.
- 04
Booked with context
Qualified candidates pick a slot in the coordinator's calendar; the full transcript attaches to the booking.
- 05
Nothing wasted
Near-misses get a courteous message and file into a database segmented for future studies.
Starts from
- Recruitment ads
- Volunteer chats
Delivers
- Eligibility scoring
- Coordinator bookings
- Answers attached
The impact
The site now screens three times more qualified candidates from the same ad spend, with volunteers pre-screened at midnight as easily as at noon. Coordinator phone-screening dropped from twenty hours a week to six, screening-visit no-shows fell by more than half, and enrollment timelines shortened on every active trial.
Questions about this automation
How does the bot decide whether a volunteer is eligible?
The bot asks the trial's inclusion and exclusion questions one at a time in plain language. OpenAI then scores the answers against the protocol criteria. Clear matches can book a screening visit, clear exclusions get a courteous message, and borderline answers are flagged for a coordinator to review before any decision is made.
What does the coordinator see when a candidate books?
The booking lands in the coordinator's calendar with the full chat transcript and a structured summary of each answer attached. The coordinator starts the screening call already knowing the candidate's responses, so the visit can focus on confirming details and explaining the study instead of repeating the basic questions.
What happens to volunteers who don't qualify for the current trial?
They receive a polite message explaining that this study is not a fit, and their answers are saved in a database segmented by key criteria. When a new study opens, the site can contact matching volunteers first, which shortens recruitment for future trials without extra ad spend.