Skip to content
Analytica

Enroll faster and stay audit-ready without adding staff.

AI systems for research sites in metro Detroit and across Michigan: first contact with a referral in seconds, prescreens run and visits booked, the protocol answered from its own pages, the regulatory binder watched, source documents read with a person confirming, and the numbers sponsors ask for in one view. Built beside the CTMS, EDC and scheduling you already run, never in their place.

Talk to us

In short

Eleven systems around a study, none of them a system of record: the referral answered before a coordinator sees it, the protocol that answers itself, the binder that warns six weeks out, the paper form read with a person confirming, and one page that shows every study against its target.

  • First contact with a referral in seconds, after hours included
  • Protocol and SOP answers with the section cited
  • Every expiring credential reminded before it lapses
  • Source documents read, every value confirmed by a person
Two researchers discussing a study on a tablet in a laboratory
Photo: Kaboompics, Pexels

Why the first call decides the study

One site in nine never enrolls a participant. The difference is usually the first call.

Recruitment is decided in the minutes after a referral, and those minutes belong to a coordinator who is in a visit.

Tufts looked at nearly 16,000 sites across 151 trials. One in nine never enrolled anyone; more than a third under-enrolled. The referrals were there. The site was in a visit when they arrived, and by morning they had answered someone else.

We build the systems around the study that nobody has hands for: the referral answered in seconds with the IRB-approved prescreen, the visit booked and reminded, the protocol answered from its own pages, the binder watched, the paper read. Each one sits beside the CTMS, the EDC and the binder, never in their place, and the study team keeps every decision that is theirs.

11%
of sites in a typical trial fail to enroll a single participant
37%
under-enroll, across nearly 16,000 sites in 151 Phase II and III trials

Tufts Center for the Study of Drug Development, Impact Report, 2013.

Eleven systems

Built around the study, never as a system of record

Each is pre-built from a site's real week and fitted to yours. Pick one to read it; the first two are where most sites start.
A man typing a message on his phone at home in the evening01 of 11

Speed-to-Lead Screening and Scheduling

First contact with an inbound referral in seconds, by text or voice, whether it came from an ad, the study website, ClinicalTrials.gov or a call. The IRB-approved prescreen for that study is run, the answers captured, clear non-fits let down kindly, and qualified candidates booked straight into the site's scheduling system with confirmation and reminders. Coordinators receive completed prescreens rather than names to chase. It works after hours, where a large share of inquiries arrive, and reports time to first contact, contact rate, prescreen completion and show rate per campaign, so recruitment spend is judged on qualified visits rather than clicks.

Works with
Study websiteClinicalTrials.govYour phone systemYour scheduling system
Never
Decides eligibility. It runs the IRB-approved questions and records the answers; the study team decides.
Photo: Ron Lach, Pexels.

Audit readiness

The credential that lapsed was found by the monitor. Now it is found six weeks earlier.

Every expiring item across every study and every person, on one board, with the reminders going out on their own.

A regulatory binder is hundreds of dated documents kept by a person who also runs visits. The CV that expired last month is not negligence; it is the sixth folder on a Friday. A monitoring visit is the wrong place to learn it.

The tracker reads every date across every study and every staff member, reminds the right person at sixty, thirty and seven days, flags what is missing per protocol, and assembles the packet when a new person or a new site is added. The binder stays the record. The board is how you know it is complete before anyone asks.

Fitted work

What comes up here, specifically

Built for how research sites in metro Detroit and across Michigan actually recruit, visit and report.
A nurse in scrubs going through a form on a tablet with a patient in a clinic waiting area
The visit happens because the reminder carried the parking, the fasting note and a reply that confirmed it. Photo: Cedric Fauntleroy, Pexels.
  1. The research corridor

    Metro Detroit runs some of the busiest trial sites in the country, the hospital systems and the independent sites around them, all drawing on the same population. Two sites can be recruiting for the same indication a few miles apart. The one that answers a referral at 9:40 pm is the one that gets the participant.

  2. The January visit

    A winter visit across Oakland or Wayne County is the one most likely to be missed. The reminder carries the parking and the weather plan, a confirmation is asked for, and the reschedule is offered inside the window before it becomes a deviation.

  3. Prescreening in Arabic and Spanish

    Dearborn, Hamtramck and southwest Detroit are under-represented in most protocols. Prescreens run in Arabic and Spanish from the IRB-approved translations, and the coordinator who follows up is told which language to use.

  4. The sponsor two time zones away

    A feasibility questionnaire that lands at 4 pm from a sponsor in California is drafted from the site profile that evening and reviewed in the morning, not started next week. Turnaround is the part of feasibility a site controls.

Getting there

Audit, pick the first system, build beside the record, verify, run

Nothing goes live because a build finished. It goes live when your QA has run the scripts and the documentation is in the quality file.
  1. 01

    Audit the site's week

    Where the referrals arrive and how long they wait, what the coordinators retype, what the binder is missing, what the sponsor's last questionnaire took. Measured, not guessed.

  2. 02

    Pick the first system

    Usually the referral response or the binder, because both show a result inside a month and neither touches a system of record.

  3. 03

    Build beside the systems of record

    On accounts the site owns, connected to the CTMS, EDC and scheduling through their own integrations, with every write logged and every script the IRB has approved used verbatim.

  4. 04

    Verify with your QA

    Test scripts run with your staff, the documentation written for your quality file, access set by role. Nothing goes live because a build finished.

  5. 05

    Run it

    Monitored so silence is an alert, repaired when a vendor changes an interface, reported monthly in the numbers a sponsor asks for.

Scope

What we do not build

Each of these is a line a QA lead or an IRB would draw, so we draw it first.
  1. A system of record

    The EDC, the CTMS and the regulatory binder stay the systems of record. What we build reads from them and writes to them only through their own integrations, with the audit trail intact.

  2. Eligibility decisions

    The prescreen runs the IRB-approved questions and records the answers. Whether someone is eligible is decided by the study team, on the protocol, at the screening visit.

  3. Consent

    No system here obtains, explains or documents consent. It books the visit at which consent happens and reminds the participant to come.

  4. Compliance certification

    We do not validate your systems or certify Part 11, GCP or HIPAA compliance. We build to the standards, document what we built, and hand the documentation to your QA.

  5. Clinical judgement

    The protocol assistant answers what the protocol says, with the section cited. It never says what to do with a participant.

  6. Training on your participants

    Nothing about a participant is used to train anything. Where a system touches identifiable information it does so under a Business Associate Agreement, in accounts the site owns.

Questions

Asked and answered

Does this touch protected health information?

Some of it does, the referral response, the visit messaging and the digitization, and those run under a Business Associate Agreement in accounts the site owns, with access by role and every action logged. The protocol assistant, the feasibility generator, the binder tracker and the request portals read documents and dates, not participants, and sit outside the regulated systems.

Is the source digitization Part 11 compliant?

It is built as a transcription accelerator with human verification, not as a validated system of record: every value is confirmed by a person against the source image before it is committed to your EDC through the EDC's own interface, and the audit trail lives there. We document the build for your quality file; validation of your systems remains yours.

Who approves the prescreen questions and the messages?

Your IRB, as they do now. The system runs the approved questions and scripts verbatim, in the approved languages, and changes to any of them go through your amendment process before they are loaded.

Does it work with our CTMS, EDC and scheduling system?

It is built beside the systems you already run, through their own integrations, and it never replaces them. We do not name vendors until we have connected one for a client; tell us what you run and we will say exactly how it connects and what it cannot do.

Who owns the site profile, the documents and the data?

You do. Everything runs in accounts the site owns, the documentation is written for whoever comes next, and if we stop working together the systems keep running where they are.

We already use a recruitment vendor. Does this replace them?

No. It answers what they send within seconds instead of the next morning, and it measures every source, the vendor included, in qualified visits rather than referrals. Most sites find the measurement changes where the recruitment budget goes.

Tell us about your enrolling studies.

How many are open, where the referrals come from, and how long the last one waited for a reply. We will tell you what the first system would be, what it would change in a month, and what your QA would need to see before it went live.

Talk to us